├── static ├── .DS_Store ├── ASR_bad.wav ├── ASR_good.wav ├── ASR_okay.wav ├── css │ ├── .DS_Store │ └── style-new.css └── javascript │ └── callbacks.js ├── data_de_en_2.pickle ├── templates ├── .DS_Store └── main.html ├── images └── MainScreen.jpg ├── utilsFileIO.py ├── requirements.txt ├── lambdaTTS.py ├── RuleBasedModels.py ├── whisper_wrapper.py ├── webApp.py ├── WordMetrics.py ├── ModelInterfaces.py ├── .gitignore ├── lambdaGetSample.py ├── AIModels.py ├── models.py ├── unitTests.py ├── README.md ├── lambdaSpeechToScore.py ├── WordMatching.py ├── pronunciationTrainer.py └── LICENSE /static/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/static/.DS_Store -------------------------------------------------------------------------------- /data_de_en_2.pickle: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/data_de_en_2.pickle -------------------------------------------------------------------------------- /static/ASR_bad.wav: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/static/ASR_bad.wav -------------------------------------------------------------------------------- /static/ASR_good.wav: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/static/ASR_good.wav -------------------------------------------------------------------------------- /static/ASR_okay.wav: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/static/ASR_okay.wav -------------------------------------------------------------------------------- /templates/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/templates/.DS_Store -------------------------------------------------------------------------------- /images/MainScreen.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/images/MainScreen.jpg -------------------------------------------------------------------------------- /static/css/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Thiagohgl/ai-pronunciation-trainer/HEAD/static/css/.DS_Store -------------------------------------------------------------------------------- /utilsFileIO.py: -------------------------------------------------------------------------------- 1 | import string 2 | import random 3 | 4 | 5 | def generateRandomString(str_length: int = 20): 6 | 7 | # printing lowercase 8 | letters = string.ascii_lowercase 9 | return ''.join(random.choice(letters) for i in range(str_length)) -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | -f https://download.pytorch.org/whl/torch_stable.html 2 | torch 3 | torchaudio 4 | soundfile 5 | omegaconf 6 | epitran 7 | audioread 8 | requests 9 | dtwalign 10 | eng_to_ipa 11 | pandas 12 | flask 13 | flask_cors 14 | pickle-mixin 15 | sqlalchemy 16 | transformers 17 | sentencepiece 18 | transformers 19 | -------------------------------------------------------------------------------- /lambdaTTS.py: -------------------------------------------------------------------------------- 1 | 2 | import models 3 | import soundfile as sf 4 | import json 5 | import AIModels 6 | #from flask import Response 7 | import utilsFileIO 8 | import os 9 | import base64 10 | 11 | sampling_rate = 16000 12 | model_TTS_lambda = AIModels.NeuralTTS(models.getTTSModel('de'), sampling_rate) 13 | 14 | 15 | def lambda_handler(event, context): 16 | 17 | body = json.loads(event['body']) 18 | 19 | text_string = body['value'] 20 | 21 | linear_factor = 0.2 22 | audio = model_TTS_lambda.getAudioFromSentence( 23 | text_string).detach().numpy()*linear_factor 24 | random_file_name = utilsFileIO.generateRandomString(20)+'.wav' 25 | 26 | sf.write('./'+random_file_name, audio, 16000) 27 | 28 | with open(random_file_name, "rb") as f: 29 | audio_byte_array = f.read() 30 | 31 | os.remove(random_file_name) 32 | 33 | 34 | return { 35 | 'statusCode': 200, 36 | 'headers': { 37 | 'Access-Control-Allow-Headers': '*', 38 | 'Access-Control-Allow-Origin': '*', 39 | 'Access-Control-Allow-Methods': 'OPTIONS,POST,GET' 40 | }, 41 | 'body': json.dumps( 42 | { 43 | "wavBase64": str(base64.b64encode(audio_byte_array))[2:-1], 44 | }, 45 | ) 46 | } 47 | -------------------------------------------------------------------------------- /RuleBasedModels.py: -------------------------------------------------------------------------------- 1 | import ModelInterfaces 2 | import torch 3 | import numpy as np 4 | import epitran 5 | import eng_to_ipa 6 | 7 | 8 | def get_phonem_converter(language: str): 9 | if language == 'de': 10 | phonem_converter = EpitranPhonemConverter( 11 | epitran.Epitran('deu-Latn')) 12 | elif language == 'en': 13 | phonem_converter = EngPhonemConverter() 14 | else: 15 | raise ValueError('Language not implemented') 16 | 17 | return phonem_converter 18 | 19 | class EpitranPhonemConverter(ModelInterfaces.ITextToPhonemModel): 20 | word_locations_in_samples = None 21 | audio_transcript = None 22 | 23 | def __init__(self, epitran_model) -> None: 24 | super().__init__() 25 | self.epitran_model = epitran_model 26 | 27 | def convertToPhonem(self, sentence: str) -> str: 28 | phonem_representation = self.epitran_model.transliterate(sentence) 29 | return phonem_representation 30 | 31 | 32 | class EngPhonemConverter(ModelInterfaces.ITextToPhonemModel): 33 | 34 | def __init__(self,) -> None: 35 | super().__init__() 36 | 37 | def convertToPhonem(self, sentence: str) -> str: 38 | phonem_representation = eng_to_ipa.convert(sentence) 39 | phonem_representation = phonem_representation.replace('*','') 40 | return phonem_representation 41 | -------------------------------------------------------------------------------- /whisper_wrapper.py: -------------------------------------------------------------------------------- 1 | import torch 2 | from transformers import pipeline 3 | from ModelInterfaces import IASRModel 4 | from typing import Union 5 | import numpy as np 6 | 7 | class WhisperASRModel(IASRModel): 8 | def __init__(self, model_name="openai/whisper-base"): 9 | self.asr = pipeline("automatic-speech-recognition", model=model_name, return_timestamps="word") 10 | self._transcript = "" 11 | self._word_locations = [] 12 | self.sample_rate = 16000 13 | 14 | def processAudio(self, audio:Union[np.ndarray, torch.Tensor]): 15 | # 'audio' can be a path to a file or a numpy array of audio samples. 16 | if isinstance(audio, torch.Tensor): 17 | audio = audio.detach().cpu().numpy() 18 | result = self.asr(audio[0]) 19 | self._transcript = result["text"] 20 | self._word_locations = [{"word": word_info["text"], 21 | "start_ts": word_info["timestamp"][0] * self.sample_rate if word_info["timestamp"][0] is not None else None, 22 | "end_ts": (word_info["timestamp"][1] * self.sample_rate if word_info["timestamp"][1] is not None else (word_info["timestamp"][0] + 1) * self.sample_rate), 23 | "tag": "processed"} for word_info in result["chunks"]] 24 | 25 | def getTranscript(self) -> str: 26 | return self._transcript 27 | 28 | def getWordLocations(self) -> list: 29 | 30 | return self._word_locations 31 | -------------------------------------------------------------------------------- /webApp.py: -------------------------------------------------------------------------------- 1 | from flask import Flask, render_template, request 2 | import webbrowser 3 | import os 4 | from flask_cors import CORS 5 | import json 6 | 7 | import lambdaTTS 8 | import lambdaSpeechToScore 9 | import lambdaGetSample 10 | 11 | app = Flask(__name__) 12 | cors = CORS(app) 13 | app.config['CORS_HEADERS'] = '*' 14 | 15 | rootPath = '' 16 | 17 | 18 | @app.route(rootPath+'/') 19 | def main(): 20 | return render_template('main.html') 21 | 22 | 23 | @app.route(rootPath+'/getAudioFromText', methods=['POST']) 24 | def getAudioFromText(): 25 | event = {'body': json.dumps(request.get_json(force=True))} 26 | return lambdaTTS.lambda_handler(event, []) 27 | 28 | 29 | @app.route(rootPath+'/getSample', methods=['POST']) 30 | def getNext(): 31 | event = {'body': json.dumps(request.get_json(force=True))} 32 | return lambdaGetSample.lambda_handler(event, []) 33 | 34 | 35 | @app.route(rootPath+'/GetAccuracyFromRecordedAudio', methods=['POST']) 36 | def GetAccuracyFromRecordedAudio(): 37 | 38 | try: 39 | event = {'body': json.dumps(request.get_json(force=True))} 40 | lambda_correct_output = lambdaSpeechToScore.lambda_handler(event, []) 41 | except Exception as e: 42 | print('Error: ', str(e)) 43 | return { 44 | 'statusCode': 200, 45 | 'headers': { 46 | 'Access-Control-Allow-Headers': '*', 47 | 'Access-Control-Allow-Credentials': "true", 48 | 'Access-Control-Allow-Origin': '*', 49 | 'Access-Control-Allow-Methods': 'OPTIONS,POST,GET' 50 | }, 51 | 'body': '' 52 | } 53 | 54 | return lambda_correct_output 55 | 56 | 57 | if __name__ == "__main__": 58 | language = 'de' 59 | print(os.system('pwd')) 60 | webbrowser.open_new('http://127.0.0.1:3000/') 61 | app.run(host="0.0.0.0", port=3000) 62 | -------------------------------------------------------------------------------- /WordMetrics.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | 3 | # ref from https://gitlab.com/-/snippets/1948157 4 | # For some variants, look here https://en.wikibooks.org/wiki/Algorithm_Implementation/Strings/Levenshtein_distance#Python 5 | 6 | # Pure python 7 | def edit_distance_python2(a, b): 8 | # This version is commutative, so as an optimization we force |a|>=|b| 9 | if len(a) < len(b): 10 | return edit_distance_python(b, a) 11 | if len(b) == 0: # Can deal with empty sequences faster 12 | return len(a) 13 | # Only two rows are really needed: the one currently filled in, and the previous 14 | distances = [] 15 | distances.append([i for i in range(len(b)+1)]) 16 | distances.append([0 for _ in range(len(b)+1)]) 17 | # We can prefill the first row: 18 | costs = [0 for _ in range(3)] 19 | for i, a_token in enumerate(a, start=1): 20 | distances[1][0] += 1 # Deals with the first column. 21 | for j, b_token in enumerate(b, start=1): 22 | costs[0] = distances[1][j-1] + 1 23 | costs[1] = distances[0][j] + 1 24 | costs[2] = distances[0][j-1] + (0 if a_token == b_token else 1) 25 | distances[1][j] = min(costs) 26 | # Move to the next row: 27 | distances[0][:] = distances[1][:] 28 | return distances[1][len(b)] 29 | 30 | #https://stackabuse.com/levenshtein-distance-and-text-similarity-in-python/ 31 | def edit_distance_python(seq1, seq2): 32 | size_x = len(seq1) + 1 33 | size_y = len(seq2) + 1 34 | matrix = np.zeros ((size_x, size_y)) 35 | for x in range(size_x): 36 | matrix [x, 0] = x 37 | for y in range(size_y): 38 | matrix [0, y] = y 39 | 40 | for x in range(1, size_x): 41 | for y in range(1, size_y): 42 | if seq1[x-1] == seq2[y-1]: 43 | matrix [x,y] = min( 44 | matrix[x-1, y] + 1, 45 | matrix[x-1, y-1], 46 | matrix[x, y-1] + 1 47 | ) 48 | else: 49 | matrix [x,y] = min( 50 | matrix[x-1,y] + 1, 51 | matrix[x-1,y-1] + 1, 52 | matrix[x,y-1] + 1 53 | ) 54 | #print (matrix) 55 | return (matrix[size_x - 1, size_y - 1]) -------------------------------------------------------------------------------- /ModelInterfaces.py: -------------------------------------------------------------------------------- 1 | 2 | import abc 3 | import numpy as np 4 | 5 | 6 | class IASRModel(metaclass=abc.ABCMeta): 7 | @classmethod 8 | def __subclasshook__(cls, subclass): 9 | return (hasattr(subclass, 'getTranscript') and 10 | callable(subclass.getTranscript) and 11 | hasattr(subclass, 'getWordLocations') and 12 | callable(subclass.getWordLocations) and 13 | hasattr(subclass, 'processAudio') and 14 | callable(subclass.processAudio)) 15 | 16 | @abc.abstractmethod 17 | def getTranscript(self) -> str: 18 | """Get the transcripts of the process audio""" 19 | raise NotImplementedError 20 | 21 | @abc.abstractmethod 22 | def getWordLocations(self) -> list: 23 | """Get the pair of words location from audio""" 24 | raise NotImplementedError 25 | 26 | @abc.abstractmethod 27 | def processAudio(self, audio): 28 | """Process the audio""" 29 | raise NotImplementedError 30 | 31 | 32 | class ITranslationModel(metaclass=abc.ABCMeta): 33 | @classmethod 34 | def __subclasshook__(cls, subclass): 35 | return (hasattr(subclass, 'translateSentence') and 36 | callable(subclass.translateSentence)) 37 | 38 | @abc.abstractmethod 39 | def translateSentence(self, str) -> str: 40 | """Get the translation of the sentence""" 41 | raise NotImplementedError 42 | 43 | 44 | class ITextToSpeechModel(metaclass=abc.ABCMeta): 45 | @classmethod 46 | def __subclasshook__(cls, subclass): 47 | return (hasattr(subclass, 'getAudioFromSentence') and 48 | callable(subclass.getAudioFromSentence)) 49 | 50 | @abc.abstractmethod 51 | def getAudioFromSentence(self, str) -> np.array: 52 | """Get audio from sentence""" 53 | raise NotImplementedError 54 | 55 | 56 | class ITextToPhonemModel(metaclass=abc.ABCMeta): 57 | @classmethod 58 | def __subclasshook__(cls, subclass): 59 | return (hasattr(subclass, 'convertToPhonem') and 60 | callable(subclass.convertToPhonem)) 61 | 62 | @abc.abstractmethod 63 | def convertToPhonem(self, str) -> str: 64 | """Convert sentence to phonemes""" 65 | raise NotImplementedError 66 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | pip-wheel-metadata/ 24 | share/python-wheels/ 25 | *.egg-info/ 26 | .installed.cfg 27 | *.egg 28 | MANIFEST 29 | 30 | # PyInstaller 31 | # Usually these files are written by a python script from a template 32 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 33 | *.manifest 34 | *.spec 35 | 36 | # Installer logs 37 | pip-log.txt 38 | pip-delete-this-directory.txt 39 | 40 | # Unit test / coverage reports 41 | htmlcov/ 42 | .tox/ 43 | .nox/ 44 | .coverage 45 | .coverage.* 46 | .cache 47 | nosetests.xml 48 | coverage.xml 49 | *.cover 50 | *.py,cover 51 | .hypothesis/ 52 | .pytest_cache/ 53 | 54 | # Translations 55 | *.mo 56 | *.pot 57 | 58 | # Django stuff: 59 | *.log 60 | local_settings.py 61 | db.sqlite3 62 | db.sqlite3-journal 63 | 64 | # Flask stuff: 65 | instance/ 66 | .webassets-cache 67 | 68 | # Scrapy stuff: 69 | .scrapy 70 | 71 | # Sphinx documentation 72 | docs/_build/ 73 | 74 | # PyBuilder 75 | target/ 76 | 77 | # Jupyter Notebook 78 | .ipynb_checkpoints 79 | 80 | # IPython 81 | profile_default/ 82 | ipython_config.py 83 | 84 | # pyenv 85 | .python-version 86 | 87 | # pipenv 88 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 89 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 90 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 91 | # install all needed dependencies. 92 | #Pipfile.lock 93 | 94 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 95 | __pypackages__/ 96 | 97 | # Celery stuff 98 | celerybeat-schedule 99 | celerybeat.pid 100 | 101 | # SageMath parsed files 102 | *.sage.py 103 | 104 | # Environments 105 | .env 106 | .venv 107 | env/ 108 | venv/ 109 | ENV/ 110 | env.bak/ 111 | venv.bak/ 112 | 113 | # Spyder project settings 114 | .spyderproject 115 | .spyproject 116 | 117 | # Rope project settings 118 | .ropeproject 119 | 120 | # mkdocs documentation 121 | /site 122 | 123 | # mypy 124 | .mypy_cache/ 125 | .dmypy.json 126 | dmypy.json 127 | 128 | # Pyre type checker 129 | .pyre/ 130 | -------------------------------------------------------------------------------- /lambdaGetSample.py: -------------------------------------------------------------------------------- 1 | 2 | import pandas as pd 3 | import json 4 | import RuleBasedModels 5 | import epitran 6 | import random 7 | import pickle 8 | 9 | 10 | class TextDataset(): 11 | def __init__(self, table): 12 | self.table_dataframe = table 13 | self.number_of_samples = len(table) 14 | 15 | def __getitem__(self, idx): 16 | 17 | line = [self.table_dataframe['sentence'].iloc[idx]] 18 | return line 19 | 20 | def __len__(self): 21 | return self.number_of_samples 22 | 23 | 24 | sample_folder = "./databases/" 25 | lambda_database = {} 26 | lambda_ipa_converter = {} 27 | available_languages = ['de', 'en'] 28 | 29 | for language in available_languages: 30 | df = pd.read_csv(sample_folder+'data_'+language+'.csv',delimiter=';') 31 | lambda_database[language] = TextDataset(df) 32 | lambda_ipa_converter[language] = RuleBasedModels.get_phonem_converter(language) 33 | 34 | lambda_translate_new_sample = False 35 | 36 | 37 | def lambda_handler(event, context): 38 | 39 | body = json.loads(event['body']) 40 | 41 | category = int(body['category']) 42 | 43 | language = body['language'] 44 | 45 | sample_in_category = False 46 | 47 | while(not sample_in_category): 48 | valid_sequence = False 49 | while not valid_sequence: 50 | try: 51 | sample_idx = random.randint(0, len(lambda_database[language])) 52 | current_transcript = lambda_database[language][ 53 | sample_idx] 54 | valid_sequence = True 55 | except: 56 | pass 57 | 58 | sentence_category = getSentenceCategory( 59 | current_transcript[0]) 60 | 61 | sample_in_category = (sentence_category == 62 | category) or category == 0 63 | 64 | translated_trascript = "" 65 | 66 | current_ipa = lambda_ipa_converter[language].convertToPhonem( 67 | current_transcript[0]) 68 | 69 | result = {'real_transcript': current_transcript, 70 | 'ipa_transcript': current_ipa, 71 | 'transcript_translation': translated_trascript} 72 | 73 | return json.dumps(result) 74 | 75 | 76 | def getSentenceCategory(sentence) -> int: 77 | number_of_words = len(sentence.split()) 78 | categories_word_limits = [0, 8, 20, 100000] 79 | for category in range(len(categories_word_limits)-1): 80 | if number_of_words > categories_word_limits[category] and number_of_words <= categories_word_limits[category+1]: 81 | return category+1 82 | -------------------------------------------------------------------------------- /AIModels.py: -------------------------------------------------------------------------------- 1 | import ModelInterfaces 2 | import torch 3 | import numpy as np 4 | 5 | 6 | class NeuralASR(ModelInterfaces.IASRModel): 7 | word_locations_in_samples = None 8 | audio_transcript = None 9 | 10 | def __init__(self, model: torch.nn.Module, decoder) -> None: 11 | super().__init__() 12 | self.model = model 13 | self.decoder = decoder # Decoder from CTC-outputs to transcripts 14 | 15 | def getTranscript(self) -> str: 16 | """Get the transcripts of the process audio""" 17 | assert(self.audio_transcript != None, 18 | 'Can get audio transcripts without having processed the audio') 19 | return self.audio_transcript 20 | 21 | def getWordLocations(self) -> list: 22 | """Get the pair of words location from audio""" 23 | assert(self.word_locations_in_samples != None, 24 | 'Can get word locations without having processed the audio') 25 | 26 | return self.word_locations_in_samples 27 | 28 | def processAudio(self, audio: torch.Tensor): 29 | """Process the audio""" 30 | audio_length_in_samples = audio.shape[1] 31 | with torch.inference_mode(): 32 | nn_output = self.model(audio) 33 | 34 | self.audio_transcript, self.word_locations_in_samples = self.decoder( 35 | nn_output[0, :, :].detach(), audio_length_in_samples, word_align=True) 36 | 37 | 38 | class NeuralTTS(ModelInterfaces.ITextToSpeechModel): 39 | def __init__(self, model: torch.nn.Module, sampling_rate: int) -> None: 40 | super().__init__() 41 | self.model = model 42 | self.sampling_rate = sampling_rate 43 | 44 | def getAudioFromSentence(self, sentence: str) -> np.array: 45 | with torch.inference_mode(): 46 | audio_transcript = self.model.apply_tts(texts=[sentence], 47 | sample_rate=self.sampling_rate)[0] 48 | 49 | return audio_transcript 50 | 51 | 52 | class NeuralTranslator(ModelInterfaces.ITranslationModel): 53 | def __init__(self, model: torch.nn.Module, tokenizer) -> None: 54 | super().__init__() 55 | self.model = model 56 | self.tokenizer = tokenizer 57 | 58 | def translateSentence(self, sentence: str) -> str: 59 | """Get the transcripts of the process audio""" 60 | tokenized_text = self.tokenizer(sentence, return_tensors='pt') 61 | translation = self.model.generate(**tokenized_text) 62 | translated_text = self.tokenizer.batch_decode( 63 | translation, skip_special_tokens=True)[0] 64 | 65 | return translated_text 66 | -------------------------------------------------------------------------------- /models.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | import pickle 4 | from ModelInterfaces import IASRModel 5 | from AIModels import NeuralASR 6 | 7 | def getASRModel(language: str,use_whisper:bool=True) -> IASRModel: 8 | 9 | if use_whisper: 10 | from whisper_wrapper import WhisperASRModel 11 | return WhisperASRModel() 12 | 13 | if language == 'de': 14 | 15 | model, decoder, utils = torch.hub.load(repo_or_dir='snakers4/silero-models', 16 | model='silero_stt', 17 | language='de', 18 | device=torch.device('cpu')) 19 | model.eval() 20 | return NeuralASR(model, decoder) 21 | 22 | elif language == 'en': 23 | model, decoder, utils = torch.hub.load(repo_or_dir='snakers4/silero-models', 24 | model='silero_stt', 25 | language='en', 26 | device=torch.device('cpu')) 27 | model.eval() 28 | return NeuralASR(model, decoder) 29 | elif language == 'fr': 30 | model, decoder, utils = torch.hub.load(repo_or_dir='snakers4/silero-models', 31 | model='silero_stt', 32 | language='fr', 33 | device=torch.device('cpu')) 34 | model.eval() 35 | return NeuralASR(model, decoder) 36 | else: 37 | raise ValueError('Language not implemented') 38 | 39 | 40 | def getTTSModel(language: str) -> nn.Module: 41 | 42 | if language == 'de': 43 | 44 | speaker = 'thorsten_v2' # 16 kHz 45 | model, _ = torch.hub.load(repo_or_dir='snakers4/silero-models', 46 | model='silero_tts', 47 | language=language, 48 | speaker=speaker) 49 | 50 | elif language == 'en': 51 | speaker = 'lj_16khz' # 16 kHz 52 | model = torch.hub.load(repo_or_dir='snakers4/silero-models', 53 | model='silero_tts', 54 | language=language, 55 | speaker=speaker) 56 | else: 57 | raise ValueError('Language not implemented') 58 | 59 | return model 60 | 61 | 62 | def getTranslationModel(language: str) -> nn.Module: 63 | from transformers import AutoTokenizer 64 | from transformers import AutoModelForSeq2SeqLM 65 | if language == 'de': 66 | model = AutoModelForSeq2SeqLM.from_pretrained( 67 | "Helsinki-NLP/opus-mt-de-en") 68 | tokenizer = AutoTokenizer.from_pretrained( 69 | "Helsinki-NLP/opus-mt-de-en") 70 | # Cache models to avoid Hugging face processing 71 | with open('translation_model_de.pickle', 'wb') as handle: 72 | pickle.dump(model, handle) 73 | with open('translation_tokenizer_de.pickle', 'wb') as handle: 74 | pickle.dump(tokenizer, handle) 75 | else: 76 | raise ValueError('Language not implemented') 77 | 78 | return model, tokenizer 79 | -------------------------------------------------------------------------------- /unitTests.py: -------------------------------------------------------------------------------- 1 | import unittest 2 | 3 | import ModelInterfaces 4 | import lambdaGetSample 5 | import RuleBasedModels 6 | import epitran 7 | import json 8 | import pronunciationTrainer 9 | 10 | 11 | def test_category(category: int, threshold_min: int, threshold_max: int): 12 | event = {'body': json.dumps({'category': category, 'language': 'de'})} 13 | for _ in range(1000): 14 | response = lambdaGetSample.lambda_handler(event, []) 15 | response_dict = json.loads(response) 16 | number_of_words = len( 17 | response_dict['real_transcript'][0].split()) 18 | length_valid = number_of_words > threshold_min and number_of_words <= threshold_max 19 | if not length_valid: 20 | print('Category ', category, 21 | ' had a sentence with length ', number_of_words) 22 | return False 23 | return True 24 | 25 | 26 | class TestDataset(unittest.TestCase): 27 | 28 | def test_random_sentences(self): 29 | 30 | self.assertFalse(test_category(0, 0, 8)) 31 | 32 | def test_easy_sentences(self): 33 | 34 | self.assertTrue(test_category(1, 0, 8)) 35 | 36 | def test_normal_sentences(self): 37 | self.assertTrue(test_category(2, 8, 20)) 38 | 39 | def test_hard_sentences(self): 40 | self.assertTrue(test_category(3, 20, 10000)) 41 | 42 | 43 | def check_phonem_converter(converter: ModelInterfaces.ITextToPhonemModel, input: str, expected_output: str): 44 | output = converter.convertToPhonem(input) 45 | 46 | is_correct = output == expected_output 47 | if not is_correct: 48 | print('Conversion from "', input, '" should be "', 49 | expected_output, '", but was "', output, '"') 50 | return is_correct 51 | 52 | 53 | class TestPhonemConverter(unittest.TestCase): 54 | 55 | def test_english(self): 56 | phonem_converter = RuleBasedModels.EngPhonemConverter() 57 | self.assertTrue(check_phonem_converter( 58 | phonem_converter, 'Hello, this is a test', 'hɛˈloʊ, ðɪs ɪz ə tɛst')) 59 | 60 | def test_german(self): 61 | phonem_converter = RuleBasedModels.EpitranPhonemConverter( 62 | epitran.Epitran('deu-Latn')) 63 | 64 | self.assertTrue(check_phonem_converter( 65 | phonem_converter, 'Hallo, das ist ein Test', 'haloː, dɑːs ɪst ain tɛst')) 66 | 67 | 68 | trainer_SST_lambda = {} 69 | trainer_SST_lambda['de'] = pronunciationTrainer.getTrainer("de") 70 | 71 | 72 | class TestScore(unittest.TestCase): 73 | 74 | def test_exact_transcription(self): 75 | words_real = 'Ich habe sehr viel glück, am leben und gesund zu sein' 76 | 77 | real_and_transcribed_words, _, _ = trainer_SST_lambda['de'].matchSampleAndRecordedWords( 78 | words_real, words_real) 79 | 80 | pronunciation_accuracy, _ = trainer_SST_lambda['de'].getPronunciationAccuracy( 81 | real_and_transcribed_words) 82 | 83 | self.assertTrue(int(pronunciation_accuracy) == 100) 84 | 85 | def test_incorrect_transcription(self): 86 | words_real = 'Ich habe sehr viel glück, am leben und gesund zu sein' 87 | words_transcribed = 'Ic hab zeh viel guck am und gesund tu sein' 88 | 89 | real_and_transcribed_words, _, _ = trainer_SST_lambda['de'].matchSampleAndRecordedWords( 90 | words_real, words_transcribed) 91 | 92 | pronunciation_accuracy, _ = trainer_SST_lambda['de'].getPronunciationAccuracy( 93 | real_and_transcribed_words) 94 | 95 | self.assertTrue(int(pronunciation_accuracy) == 71) 96 | 97 | 98 | if __name__ == '__main__': 99 | unittest.main() 100 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | 2 | # AI Pronunciation Trainer 3 | This tool uses AI to evaluate your pronunciation so you can improve it and be understood more clearly. You can go straight test the tool at https://aipronunciationtr.com (please use the chrome browser for desktop and have some patience for it to "warm-up" :) ). 4 | 5 | ![](images/MainScreen.jpg) 6 | 7 | ## !! Update !! 8 | 9 | I'm happy (and surprised!) that apparently a considerable amount of people are using this tool, and this made -even more- clear that many things could be improved. I tried to solve the most critical issues with small changes on the code base, so that now: 10 | * Requirements are relaxed and should work with modern python versions (tested with 3.12) 11 | * Whisper is used as standard ASR model 12 | * You can edit the text you want to speak by clicking on it and typing (please do before an evaluation) 13 | * Code does not dependent on OR-Tools anymore, which makes is somewhat lighter and faster 14 | * Code is slightly changed so it is now easier to add more languages and change the database. 15 | * Errors will be shown on the console for easier debugging. 16 | 17 | I hope this facilitates the use of the tool and you have more fun (and learning!) with it. I plan to add those changes to the server in the coming weeks when I find some more time. 18 | 19 | ## Installation 20 | To run the program locally, you need to install the requirements and run the main python file: 21 | ``` 22 | pip install -r requirements.txt 23 | python webApp.py 24 | ``` 25 | You'll also need ffmpeg, which you can download from here https://ffmpeg.org/download.html. On Windows, it may be needed to add the ffmpeg "bin" folder to your PATH environment variable. On Mac, you can also just run "brew install ffmpeg". 26 | 27 | You should be able to run it locally without any major issues as long as you’re using a recent python 3.X version. 28 | 29 | ## Online version 30 | For the people who don’t feel comfortable running code or just want to have a quick way to use the tool, I hosted an online version of it at https://aipronunciationtr.com. It should work well in desktop-chrome, any other browser is not officially supported, although most of the functionality should work fine. 31 | 32 | Please be aware that the usage is limited by day (I’m still not rich ;)). If, for some reason, you would like to avoid the daily usage limit, just enter in contact and we see what we can do. 33 | 34 | ## Motivation 35 | 36 | Often, when we want to improve our pronunciation, it is very difficult to self-assess how good we’re speaking. Asking a native, or language instructor, to constantly correct us is either impractical, due to monetary constrains, or annoying due to simply being too boring for this other person. Additionally, they may often say “it sounds good” after your 10th try to not discourage you, even though you may still have some mistakes in your pronunciation. 37 | 38 | The AI pronunciation trainer is a way to provide objective feedback on how well your pronunciation is in an automatic and scalable fashion, so the only limit to your improvement is your own dedication. 39 | 40 | This project originated from a small program that I did to improve my own pronunciation. When I finished it, I believed it could be a useful tool also for other people trying to be better understood, so I decided to make a simple, more user-friendly version of it. 41 | 42 | ## Disclaimer 43 | This is a simple project that I made in my free time with the goal to be useful to some people. It is not perfect, thus be aware that some small bugs may be present. In case you find something is not working, all feedback is welcome, and issues may be addressed depending on their severity. 44 | 45 | ## FAQ 46 | 47 | ### How do I add a new language? 48 | 49 | There's definitely a code architecture that would allow this to be easier, but I have limited time and I think the current solution is doable with relative ease. What you need to do is: 50 | #### Backend 51 | As long as your language is supported by Whisper, you need only a database and small changes in key files: 52 | 53 | 1. Add your language identifier to the "lambdaGetSample.py" file 54 | 2. Add a .csv file with your text database in the "./databases" folder, following the naming convention 55 | 3. Add a corresponding phonem model in the "RuleBasedModels.py", you likely need only to give the language code to Epitran and possibly correct some characters with a wrapper 56 | 4. Add your trainer to "lambdaSpeechToScore.py" with the correct language code 57 | 58 | If you language is not supported by Whisper, you need to have an Speech-To-Text model and add it to the "getASRModel" function in "models.py", and it needs to implement the "IASRModel" interface. Besides this, you need to do the same as above. 59 | #### Frontend 60 | 61 | 1. In the "callback.js" function, add a case for your language 62 | 2. In the "main.html", add your language in the "languageBox" with corresponding call to the javascript function. -------------------------------------------------------------------------------- /lambdaSpeechToScore.py: -------------------------------------------------------------------------------- 1 | 2 | import torch 3 | import json 4 | import os 5 | import WordMatching as wm 6 | import utilsFileIO 7 | import pronunciationTrainer 8 | import base64 9 | import time 10 | import audioread 11 | import numpy as np 12 | from torchaudio.transforms import Resample 13 | import io 14 | import tempfile 15 | 16 | trainer_SST_lambda = {} 17 | trainer_SST_lambda['de'] = pronunciationTrainer.getTrainer("de") 18 | trainer_SST_lambda['en'] = pronunciationTrainer.getTrainer("en") 19 | 20 | transform = Resample(orig_freq=48000, new_freq=16000) 21 | 22 | 23 | def lambda_handler(event, context): 24 | 25 | data = json.loads(event['body']) 26 | 27 | real_text = data['title'] 28 | file_bytes = base64.b64decode( 29 | data['base64Audio'][22:].encode('utf-8')) 30 | language = data['language'] 31 | 32 | if len(real_text) == 0: 33 | return { 34 | 'statusCode': 200, 35 | 'headers': { 36 | 'Access-Control-Allow-Headers': '*', 37 | 'Access-Control-Allow-Credentials': "true", 38 | 'Access-Control-Allow-Origin': '*', 39 | 'Access-Control-Allow-Methods': 'OPTIONS,POST,GET' 40 | }, 41 | 'body': '' 42 | } 43 | 44 | tmp = tempfile.NamedTemporaryFile(suffix=".ogg", delete=False) 45 | tmp_name = tmp.name 46 | 47 | try: 48 | tmp.write(file_bytes) 49 | tmp.flush() 50 | 51 | tmp.close() 52 | 53 | signal, fs = audioread_load(tmp_name) 54 | 55 | finally: 56 | 57 | os.remove(tmp_name) 58 | 59 | signal = transform(torch.Tensor(signal)).unsqueeze(0) 60 | 61 | result = trainer_SST_lambda[language].processAudioForGivenText( 62 | signal, real_text) 63 | 64 | start = time.time() 65 | real_transcripts_ipa = ' '.join( 66 | [word[0] for word in result['real_and_transcribed_words_ipa']]) 67 | matched_transcripts_ipa = ' '.join( 68 | [word[1] for word in result['real_and_transcribed_words_ipa']]) 69 | 70 | real_transcripts = ' '.join( 71 | [word[0] for word in result['real_and_transcribed_words']]) 72 | matched_transcripts = ' '.join( 73 | [word[1] for word in result['real_and_transcribed_words']]) 74 | 75 | words_real = real_transcripts.lower().split() 76 | mapped_words = matched_transcripts.split() 77 | 78 | is_letter_correct_all_words = '' 79 | for idx, word_real in enumerate(words_real): 80 | 81 | mapped_letters, mapped_letters_indices = wm.get_best_mapped_words( 82 | mapped_words[idx], word_real) 83 | 84 | is_letter_correct = wm.getWhichLettersWereTranscribedCorrectly( 85 | word_real, mapped_letters) 86 | 87 | is_letter_correct_all_words += ''.join([str(is_correct) 88 | for is_correct in is_letter_correct]) + ' ' 89 | 90 | pair_accuracy_category = ' '.join( 91 | [str(category) for category in result['pronunciation_categories']]) 92 | print('Time to post-process results: ', str(time.time()-start)) 93 | 94 | res = {'real_transcript': result['recording_transcript'], 95 | 'ipa_transcript': result['recording_ipa'], 96 | 'pronunciation_accuracy': str(int(result['pronunciation_accuracy'])), 97 | 'real_transcripts': real_transcripts, 'matched_transcripts': matched_transcripts, 98 | 'real_transcripts_ipa': real_transcripts_ipa, 'matched_transcripts_ipa': matched_transcripts_ipa, 99 | 'pair_accuracy_category': pair_accuracy_category, 100 | 'start_time': result['start_time'], 101 | 'end_time': result['end_time'], 102 | 'is_letter_correct_all_words': is_letter_correct_all_words} 103 | 104 | return json.dumps(res) 105 | 106 | 107 | 108 | 109 | def audioread_load(path, offset=0.0, duration=None, dtype=np.float32): 110 | """Load an audio buffer using audioread. 111 | 112 | This loads one block at a time, and then concatenates the results. 113 | """ 114 | 115 | y = [] 116 | with audioread.audio_open(path) as input_file: 117 | sr_native = input_file.samplerate 118 | n_channels = input_file.channels 119 | 120 | s_start = int(np.round(sr_native * offset)) * n_channels 121 | 122 | if duration is None: 123 | s_end = np.inf 124 | else: 125 | s_end = s_start + \ 126 | (int(np.round(sr_native * duration)) * n_channels) 127 | 128 | n = 0 129 | 130 | for frame in input_file: 131 | frame = buf_to_float(frame, dtype=dtype) 132 | n_prev = n 133 | n = n + len(frame) 134 | 135 | if n < s_start: 136 | # offset is after the current frame 137 | # keep reading 138 | continue 139 | 140 | if s_end < n_prev: 141 | # we're off the end. stop reading 142 | break 143 | 144 | if s_end < n: 145 | # the end is in this frame. crop. 146 | frame = frame[: s_end - n_prev] 147 | 148 | if n_prev <= s_start <= n: 149 | # beginning is in this frame 150 | frame = frame[(s_start - n_prev):] 151 | 152 | # tack on the current frame 153 | y.append(frame) 154 | 155 | if y: 156 | y = np.concatenate(y) 157 | if n_channels > 1: 158 | y = y.reshape((-1, n_channels)).T 159 | else: 160 | y = np.empty(0, dtype=dtype) 161 | 162 | return y, sr_native 163 | 164 | # From Librosa 165 | 166 | 167 | def buf_to_float(x, n_bytes=2, dtype=np.float32): 168 | """Convert an integer buffer to floating point values. 169 | This is primarily useful when loading integer-valued wav data 170 | into numpy arrays. 171 | 172 | Parameters 173 | ---------- 174 | x : np.ndarray [dtype=int] 175 | The integer-valued data buffer 176 | 177 | n_bytes : int [1, 2, 4] 178 | The number of bytes per sample in ``x`` 179 | 180 | dtype : numeric type 181 | The target output type (default: 32-bit float) 182 | 183 | Returns 184 | ------- 185 | x_float : np.ndarray [dtype=float] 186 | The input data buffer cast to floating point 187 | """ 188 | 189 | # Invert the scale of the data 190 | scale = 1.0 / float(1 << ((8 * n_bytes) - 1)) 191 | 192 | # Construct the format string 193 | fmt = " np.ndarray: 14 | number_of_real_words = len(words_real) 15 | number_of_estimated_words = len(words_estimated) 16 | 17 | word_distance_matrix = np.zeros( 18 | (number_of_estimated_words+offset_blank, number_of_real_words)) 19 | for idx_estimated in range(number_of_estimated_words): 20 | for idx_real in range(number_of_real_words): 21 | word_distance_matrix[idx_estimated, idx_real] = WordMetrics.edit_distance_python( 22 | words_estimated[idx_estimated], words_real[idx_real]) 23 | 24 | if offset_blank == 1: 25 | for idx_real in range(number_of_real_words): 26 | word_distance_matrix[number_of_estimated_words, 27 | idx_real] = len(words_real[idx_real]) 28 | return word_distance_matrix 29 | 30 | 31 | def get_best_path_from_distance_matrix(word_distance_matrix): 32 | modelCpp = cp_model.CpModel() 33 | 34 | number_of_real_words = word_distance_matrix.shape[1] 35 | number_of_estimated_words = word_distance_matrix.shape[0]-1 36 | 37 | number_words = np.maximum(number_of_real_words, number_of_estimated_words) 38 | 39 | estimated_words_order = [modelCpp.NewIntVar(0, int( 40 | number_words - 1 + offset_blank), 'w%i' % i) for i in range(number_words+offset_blank)] 41 | 42 | # They are in ascending order 43 | for word_idx in range(number_words-1): 44 | modelCpp.Add( 45 | estimated_words_order[word_idx+1] >= estimated_words_order[word_idx]) 46 | 47 | total_phoneme_distance = 0 48 | real_word_at_time = {} 49 | for idx_estimated in range(number_of_estimated_words): 50 | for idx_real in range(number_of_real_words): 51 | real_word_at_time[idx_estimated, idx_real] = modelCpp.NewBoolVar( 52 | 'real_word_at_time'+str(idx_real)+'-'+str(idx_estimated)) 53 | modelCpp.Add(estimated_words_order[idx_estimated] == idx_real).OnlyEnforceIf( 54 | real_word_at_time[idx_estimated, idx_real]) 55 | total_phoneme_distance += word_distance_matrix[idx_estimated, 56 | idx_real]*real_word_at_time[idx_estimated, idx_real] 57 | 58 | # If no word in time, difference is calculated from empty string 59 | for idx_real in range(number_of_real_words): 60 | word_has_a_match = modelCpp.NewBoolVar( 61 | 'word_has_a_match'+str(idx_real)) 62 | modelCpp.Add(sum([real_word_at_time[idx_estimated, idx_real] for idx_estimated in range( 63 | number_of_estimated_words)]) == 1).OnlyEnforceIf(word_has_a_match) 64 | total_phoneme_distance += word_distance_matrix[number_of_estimated_words, 65 | idx_real]*word_has_a_match.Not() 66 | 67 | # Loss should be minimized 68 | modelCpp.Minimize(total_phoneme_distance) 69 | 70 | solver = cp_model.CpSolver() 71 | solver.parameters.max_time_in_seconds = TIME_THRESHOLD_MAPPING 72 | status = solver.Solve(modelCpp) 73 | 74 | mapped_indices = [] 75 | try: 76 | for word_idx in range(number_words): 77 | mapped_indices.append( 78 | (solver.Value(estimated_words_order[word_idx]))) 79 | 80 | return np.array(mapped_indices, dtype=int) 81 | except: 82 | return [] 83 | 84 | 85 | def get_resulting_string(mapped_indices: np.ndarray, words_estimated: list, words_real: list) -> Tuple[List,List]: 86 | mapped_words = [] 87 | mapped_words_indices = [] 88 | WORD_NOT_FOUND_TOKEN = '-' 89 | number_of_real_words = len(words_real) 90 | for word_idx in range(number_of_real_words): 91 | position_of_real_word_indices = np.where( 92 | mapped_indices == word_idx)[0].astype(int) 93 | 94 | if len(position_of_real_word_indices) == 0: 95 | mapped_words.append(WORD_NOT_FOUND_TOKEN) 96 | mapped_words_indices.append(-1) 97 | continue 98 | 99 | if len(position_of_real_word_indices) == 1: 100 | mapped_words.append( 101 | words_estimated[position_of_real_word_indices[0]]) 102 | mapped_words_indices.append(position_of_real_word_indices[0]) 103 | continue 104 | # Check which index gives the lowest error 105 | if len(position_of_real_word_indices) > 1: 106 | error = 99999 107 | best_possible_combination = '' 108 | best_possible_idx = -1 109 | for single_word_idx in position_of_real_word_indices: 110 | idx_above_word = single_word_idx >= len(words_estimated) 111 | if idx_above_word: 112 | continue 113 | error_word = WordMetrics.edit_distance_python( 114 | words_estimated[single_word_idx], words_real[word_idx]) 115 | if error_word < error: 116 | error = error_word*1 117 | best_possible_combination = words_estimated[single_word_idx] 118 | best_possible_idx = single_word_idx 119 | 120 | mapped_words.append(best_possible_combination) 121 | mapped_words_indices.append(best_possible_idx) 122 | continue 123 | 124 | return mapped_words, mapped_words_indices 125 | 126 | 127 | def get_best_mapped_words(words_estimated: list, words_real: list,use_dtw:bool = True) -> list: 128 | 129 | word_distance_matrix = get_word_distance_matrix( 130 | words_estimated, words_real) 131 | 132 | start = time.time() 133 | 134 | if use_dtw: 135 | alignment = (dtw_from_distance_matrix( 136 | word_distance_matrix.T)) 137 | 138 | mapped_indices = alignment.get_warping_path()[:len(words_estimated)] 139 | duration_of_mapping = time.time()-start 140 | else: 141 | mapped_indices = get_best_path_from_distance_matrix(word_distance_matrix) 142 | 143 | duration_of_mapping = time.time()-start 144 | # In case or-tools doesn't converge, go to a faster, low-quality solution 145 | if len(mapped_indices) == 0 or duration_of_mapping > TIME_THRESHOLD_MAPPING+0.5: 146 | #mapped_indices = (dtw_from_distance_matrix( 147 | # word_distance_matrix)).path[:len(words_estimated), 1] 148 | alignment = (dtw_from_distance_matrix( 149 | word_distance_matrix.T)) 150 | 151 | mapped_indices = alignment.get_warping_path() 152 | 153 | mapped_words, mapped_words_indices = get_resulting_string( 154 | mapped_indices, words_estimated, words_real) 155 | 156 | return mapped_words, mapped_words_indices 157 | 158 | 159 | # Faster, but not optimal 160 | def get_best_mapped_words_dtw(words_estimated: list, words_real: list) -> list: 161 | 162 | from dtwalign import dtw_from_distance_matrix 163 | word_distance_matrix = get_word_distance_matrix( 164 | words_estimated, words_real) 165 | mapped_indices = dtw_from_distance_matrix( 166 | word_distance_matrix).path[:-1, 0] 167 | 168 | mapped_words, mapped_words_indices = get_resulting_string( 169 | mapped_indices, words_estimated, words_real) 170 | return mapped_words, mapped_words_indices 171 | 172 | 173 | def getWhichLettersWereTranscribedCorrectly(real_word, transcribed_word): 174 | is_leter_correct = [None]*len(real_word) 175 | for idx, letter in enumerate(real_word): 176 | letter = letter.lower() 177 | transcribed_word[idx] = transcribed_word[idx].lower() 178 | if letter == transcribed_word[idx] or letter in punctuation: 179 | is_leter_correct[idx] = 1 180 | else: 181 | is_leter_correct[idx] = 0 182 | return is_leter_correct 183 | 184 | 185 | def parseLetterErrorsToHTML(word_real, is_leter_correct): 186 | word_colored = '' 187 | correct_color_start = '*' 188 | correct_color_end = '*' 189 | wrong_color_start = '-' 190 | wrong_color_end = '-' 191 | for idx, letter in enumerate(word_real): 192 | if is_leter_correct[idx] == 1: 193 | word_colored += correct_color_start + letter+correct_color_end 194 | else: 195 | word_colored += wrong_color_start + letter+wrong_color_end 196 | return word_colored 197 | -------------------------------------------------------------------------------- /templates/main.html: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | AI pronunciation trainer 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 |
29 | 30 |
31 | 32 | 33 |
34 | home 35 |
36 | 37 |

AI Pronunciation Trainer 38 |

39 |
40 | 41 | 42 |
43 |
44 | 45 | 46 |

47 | Language:

48 | 49 |
50 | 52 |
53 | German 55 | English 57 |
58 |
59 | 60 |

62 | | Score: 0 63 |

64 | 65 |
66 |
67 | 68 |
69 | 70 |
71 | 72 | 73 |
74 | 75 |
76 | 77 | play_arrow 79 | 80 | 81 | record_voice_over 84 | 85 |

87 | - 88 |

89 | 90 |
91 | 92 |
93 | 94 |

Click on the bar on the 95 | right 96 | to 97 | generate a 98 | new sentence (please use chrome web browser). 99 |

100 |

Before speaking, click on the mic button 101 | below to start recording and then click again when you're done. 102 |

103 |

On the left bottom you can choose the 104 | difficult. On the upper left you can choose the language. 105 |

106 |

The corresponding IPA reading of each 107 | sentence will also be displayed. If you never heard from IPA, you can check out this 108 | playlist. Try to get at least 690 points a day. Don't be shy! You can do it 110 | :) 111 |

112 | 113 |
114 | 115 |
116 |
119 |
120 | 121 | 122 | 123 | 124 |
126 |

Reference | Spoken 128 |

129 |
130 | 131 | 132 |
133 | mic 135 | 136 |
137 | 138 | 139 |
140 | 141 | 142 | 143 | 144 |
145 | 146 |
147 | 148 | 149 |

By Thiago 150 | Lobato.

151 | 152 |
153 | 155 | 156 | 158 | 159 | 160 | 161 | 163 | 164 | 166 | 167 | 168 | 169 |
170 | 171 | 172 | 173 | 174 | 175 | 176 | 177 | 178 | 179 | 180 | 181 | -------------------------------------------------------------------------------- /pronunciationTrainer.py: -------------------------------------------------------------------------------- 1 | 2 | import torch 3 | import numpy as np 4 | import models as mo 5 | import WordMetrics 6 | import WordMatching as wm 7 | import epitran 8 | import ModelInterfaces as mi 9 | import AIModels 10 | import RuleBasedModels 11 | from string import punctuation 12 | import time 13 | 14 | 15 | def getTrainer(language: str): 16 | 17 | asr_model = mo.getASRModel(language,use_whisper=True) 18 | 19 | if language == 'de': 20 | phonem_converter = RuleBasedModels.EpitranPhonemConverter( 21 | epitran.Epitran('deu-Latn')) 22 | elif language == 'en': 23 | phonem_converter = RuleBasedModels.EngPhonemConverter() 24 | else: 25 | raise ValueError('Language not implemented') 26 | 27 | trainer = PronunciationTrainer( 28 | asr_model, phonem_converter) 29 | 30 | return trainer 31 | 32 | 33 | class PronunciationTrainer: 34 | current_transcript: str 35 | current_ipa: str 36 | 37 | current_recorded_audio: torch.Tensor 38 | current_recorded_transcript: str 39 | current_recorded_word_locations: list 40 | current_recorded_intonations: torch.tensor 41 | current_words_pronunciation_accuracy = [] 42 | categories_thresholds = np.array([80, 60, 59]) 43 | 44 | sampling_rate = 16000 45 | 46 | def __init__(self, asr_model: mi.IASRModel, word_to_ipa_coverter: mi.ITextToPhonemModel) -> None: 47 | self.asr_model = asr_model 48 | self.ipa_converter = word_to_ipa_coverter 49 | 50 | def getTranscriptAndWordsLocations(self, audio_length_in_samples: int): 51 | 52 | audio_transcript = self.asr_model.getTranscript() 53 | word_locations_in_samples = self.asr_model.getWordLocations() 54 | 55 | fade_duration_in_samples = 0.05*self.sampling_rate 56 | word_locations_in_samples = [(int(np.maximum(0, word['start_ts']-fade_duration_in_samples)), int(np.minimum( 57 | audio_length_in_samples-1, word['end_ts']+fade_duration_in_samples))) for word in word_locations_in_samples] 58 | 59 | return audio_transcript, word_locations_in_samples 60 | 61 | def getWordsRelativeIntonation(self, Audio: torch.tensor, word_locations: list): 62 | intonations = torch.zeros((len(word_locations), 1)) 63 | intonation_fade_samples = 0.3*self.sampling_rate 64 | print(intonations.shape) 65 | for word in range(len(word_locations)): 66 | intonation_start = int(np.maximum( 67 | 0, word_locations[word][0]-intonation_fade_samples)) 68 | intonation_end = int(np.minimum( 69 | Audio.shape[1]-1, word_locations[word][1]+intonation_fade_samples)) 70 | intonations[word] = torch.sqrt(torch.mean( 71 | Audio[0][intonation_start:intonation_end]**2)) 72 | 73 | intonations = intonations/torch.mean(intonations) 74 | return intonations 75 | 76 | ##################### ASR Functions ########################### 77 | 78 | def processAudioForGivenText(self, recordedAudio: torch.Tensor = None, real_text=None): 79 | 80 | start = time.time() 81 | recording_transcript, recording_ipa, word_locations = self.getAudioTranscript( 82 | recordedAudio) 83 | print('Time for NN to transcript audio: ', str(time.time()-start)) 84 | 85 | start = time.time() 86 | real_and_transcribed_words, real_and_transcribed_words_ipa, mapped_words_indices = self.matchSampleAndRecordedWords( 87 | real_text, recording_transcript) 88 | print('Time for matching transcripts: ', str(time.time()-start)) 89 | 90 | start_time, end_time = self.getWordLocationsFromRecordInSeconds( 91 | word_locations, mapped_words_indices) 92 | 93 | pronunciation_accuracy, current_words_pronunciation_accuracy = self.getPronunciationAccuracy( 94 | real_and_transcribed_words) # _ipa 95 | 96 | pronunciation_categories = self.getWordsPronunciationCategory( 97 | current_words_pronunciation_accuracy) 98 | 99 | result = {'recording_transcript': recording_transcript, 100 | 'real_and_transcribed_words': real_and_transcribed_words, 101 | 'recording_ipa': recording_ipa, 'start_time': start_time, 'end_time': end_time, 102 | 'real_and_transcribed_words_ipa': real_and_transcribed_words_ipa, 'pronunciation_accuracy': pronunciation_accuracy, 103 | 'pronunciation_categories': pronunciation_categories} 104 | 105 | return result 106 | 107 | def getAudioTranscript(self, recordedAudio: torch.Tensor = None): 108 | current_recorded_audio = recordedAudio 109 | 110 | current_recorded_audio = self.preprocessAudio( 111 | current_recorded_audio) 112 | 113 | self.asr_model.processAudio(current_recorded_audio) 114 | 115 | current_recorded_transcript, current_recorded_word_locations = self.getTranscriptAndWordsLocations( 116 | current_recorded_audio.shape[1]) 117 | current_recorded_ipa = self.ipa_converter.convertToPhonem( 118 | current_recorded_transcript) 119 | 120 | return current_recorded_transcript, current_recorded_ipa, current_recorded_word_locations 121 | 122 | def getWordLocationsFromRecordInSeconds(self, word_locations, mapped_words_indices) -> list: 123 | start_time = [] 124 | end_time = [] 125 | for word_idx in range(len(mapped_words_indices)): 126 | start_time.append(float(word_locations[mapped_words_indices[word_idx]] 127 | [0])/self.sampling_rate) 128 | end_time.append(float(word_locations[mapped_words_indices[word_idx]] 129 | [1])/self.sampling_rate) 130 | return ' '.join([str(time) for time in start_time]), ' '.join([str(time) for time in end_time]) 131 | 132 | ##################### END ASR Functions ########################### 133 | 134 | ##################### Evaluation Functions ########################### 135 | def matchSampleAndRecordedWords(self, real_text, recorded_transcript): 136 | words_estimated = recorded_transcript.split() 137 | 138 | if real_text is None: 139 | words_real = self.current_transcript[0].split() 140 | else: 141 | words_real = real_text.split() 142 | 143 | mapped_words, mapped_words_indices = wm.get_best_mapped_words( 144 | words_estimated, words_real) 145 | 146 | real_and_transcribed_words = [] 147 | real_and_transcribed_words_ipa = [] 148 | for word_idx in range(len(words_real)): 149 | if word_idx >= len(mapped_words)-1: 150 | mapped_words.append('-') 151 | real_and_transcribed_words.append( 152 | (words_real[word_idx], mapped_words[word_idx])) 153 | real_and_transcribed_words_ipa.append((self.ipa_converter.convertToPhonem(words_real[word_idx]), 154 | self.ipa_converter.convertToPhonem(mapped_words[word_idx]))) 155 | return real_and_transcribed_words, real_and_transcribed_words_ipa, mapped_words_indices 156 | 157 | def getPronunciationAccuracy(self, real_and_transcribed_words_ipa) -> float: 158 | total_mismatches = 0. 159 | number_of_phonemes = 0. 160 | current_words_pronunciation_accuracy = [] 161 | for pair in real_and_transcribed_words_ipa: 162 | 163 | real_without_punctuation = self.removePunctuation(pair[0]).lower() 164 | number_of_word_mismatches = WordMetrics.edit_distance_python( 165 | real_without_punctuation, self.removePunctuation(pair[1]).lower()) 166 | total_mismatches += number_of_word_mismatches 167 | number_of_phonemes_in_word = len(real_without_punctuation) 168 | number_of_phonemes += number_of_phonemes_in_word 169 | 170 | current_words_pronunciation_accuracy.append(float( 171 | number_of_phonemes_in_word-number_of_word_mismatches)/number_of_phonemes_in_word*100) 172 | 173 | percentage_of_correct_pronunciations = ( 174 | number_of_phonemes-total_mismatches)/number_of_phonemes*100 175 | 176 | return np.round(percentage_of_correct_pronunciations), current_words_pronunciation_accuracy 177 | 178 | def removePunctuation(self, word: str) -> str: 179 | return ''.join([char for char in word if char not in punctuation]) 180 | 181 | def getWordsPronunciationCategory(self, accuracies) -> list: 182 | categories = [] 183 | 184 | for accuracy in accuracies: 185 | categories.append( 186 | self.getPronunciationCategoryFromAccuracy(accuracy)) 187 | 188 | return categories 189 | 190 | def getPronunciationCategoryFromAccuracy(self, accuracy) -> int: 191 | return np.argmin(abs(self.categories_thresholds-accuracy)) 192 | 193 | def preprocessAudio(self, audio: torch.tensor) -> torch.tensor: 194 | audio = audio-torch.mean(audio) 195 | audio = audio/torch.max(torch.abs(audio)) 196 | return audio 197 | -------------------------------------------------------------------------------- /static/css/style-new.css: -------------------------------------------------------------------------------- 1 | body { 2 | background: #f2f2f2; 3 | } 4 | 5 | 6 | .expanded { 7 | margin: auto; 8 | align-content: center; 9 | } 10 | 11 | p { 12 | overflow: auto; 13 | } 14 | 15 | h1 { 16 | margin-left: 2%; 17 | } 18 | 19 | a.disabled { 20 | pointer-events: none; 21 | color: #ccc; 22 | background-color: #ccc; 23 | } 24 | 25 | 26 | .horizontal-flexbox { 27 | height: 100%; 28 | width: 100%; 29 | display: flex; 30 | } 31 | 32 | /* ############## Next button ##### */ 33 | .button-next { 34 | border-radius: 4px; 35 | display: block; 36 | border: none; 37 | color: #FFFFFF; 38 | text-align: left; 39 | font-size: 3em; 40 | box-sizing: border-box; 41 | position: absolute; 42 | top: 0; 43 | left: 0%; 44 | right: 2%; 45 | bottom: 2%; 46 | background-color: #58636d; 47 | width: 10em; 48 | 49 | transition: all 0.5s; 50 | cursor: pointer; 51 | } 52 | 53 | .button-next:hover { 54 | background-color: #6383a1 !important; 55 | } 56 | 57 | .button-next span { 58 | cursor: pointer; 59 | display: inline-block; 60 | position: relative; 61 | transition: 0.5s; 62 | } 63 | 64 | /* 65 | .button-next span:after { 66 | content: '\00bb'; 67 | position: absolute; 68 | opacity: 0; 69 | top: 0; 70 | right: -20px; 71 | transition: 0.5s; 72 | 73 | }*/ 74 | 75 | .button-next:hover span { 76 | padding-right: 25px; 77 | } 78 | 79 | .button-next:hover span:after { 80 | opacity: 1; 81 | right: 0; 82 | } 83 | 84 | 85 | 86 | /* ############# Texts ############## */ 87 | 88 | .main-text { 89 | font-size: 2.5em; 90 | max-width: 87%; 91 | } 92 | 93 | .ipa-text { 94 | font-size: 1.8em; 95 | max-width: 87%; 96 | } 97 | 98 | .ipa-text-small { 99 | font-size: 1.5em; 100 | } 101 | 102 | .accuracy-text { 103 | /*font-family: "Dank Mono", ui-monospace, monospace;*/ 104 | background: linear-gradient(to right, 105 | rgb(54, 56, 80), 106 | rgb(21, 60, 87)); 107 | background-clip: text; 108 | -webkit-background-clip: text; 109 | -webkit-text-fill-color: transparent; 110 | text-align: center; 111 | font-size: 2em; 112 | margin-left: 2%; 113 | left: 0%; 114 | } 115 | 116 | .main-text-div { 117 | overflow-y: auto; 118 | position: absolute; 119 | left: 10%; 120 | right: 10%; 121 | top: 2%; 122 | bottom: 2%; 123 | } 124 | 125 | /* ############# Card Container ############## */ 126 | .container { 127 | display: block; 128 | position: absolute; 129 | left: 2%; 130 | top: 18%; 131 | transform: translate(-0%, -0%); 132 | height: 59%; 133 | width: 96%; 134 | max-width: 96%; 135 | background: #ffff; 136 | overflow: hidden; 137 | border-radius: 20px; 138 | box-shadow: 0 0 20px 8px #d0d0d0; 139 | } 140 | 141 | .container-small { 142 | position: fixed; 143 | left: 68%; 144 | top: 79%; 145 | transform: translate(-0%, -0%); 146 | height: 7%; 147 | width: 30%; 148 | background: #ffff; 149 | overflow: hidden; 150 | border-radius: 20px; 151 | box-shadow: 0 0 20px 8px #d0d0d0; 152 | } 153 | 154 | /* ############# Icon Button ############## */ 155 | 156 | .round-button { 157 | box-sizing: border-box; 158 | display: block; 159 | width: 3em; 160 | /* 80px */ 161 | height: 3em; 162 | left: 0%; 163 | padding-top: 14px; 164 | padding-left: 0px; 165 | line-height: 0px; 166 | border: 6px solid #fff; 167 | border-radius: 50%; 168 | color: #f5f5f5; 169 | text-align: center; 170 | text-decoration: none; 171 | background-color: #467387; 172 | font-size: 20px; 173 | font-weight: bold; 174 | transition: all 0.3s ease; 175 | } 176 | 177 | .round-button:hover { 178 | background-color: rgba(0, 0, 0, 0.8); 179 | box-shadow: 0px 0px 10px #61a4d4; 180 | text-shadow: 0px 0px 10px #61a4d4; 181 | } 182 | 183 | .icon-text { 184 | font-size: 1em !important; 185 | text-align: center; 186 | } 187 | 188 | .round-button-mic { 189 | box-sizing: border-box; 190 | display: block; 191 | width: 4.5em; 192 | /* 80px */ 193 | height: 4.5em; 194 | padding-top: 14px; 195 | padding-left: -2.25em; 196 | line-height: 0px; 197 | border: 6px solid #fff; 198 | border-radius: 50%; 199 | color: #f5f5f5; 200 | text-align: center; 201 | text-decoration: none; 202 | background-color: #49d67d; 203 | /*#467387;*/ 204 | font-size: 20px; 205 | font-weight: bold; 206 | transition: all 0.3s ease; 207 | } 208 | 209 | .round-button-mic:hover { 210 | background-color: #477c5b; 211 | /*rgba(0,0,0,0.8);*/ 212 | box-shadow: 0px 0px 10px #61a4d4; 213 | text-shadow: 0px 0px 10px #61a4d4; 214 | } 215 | 216 | .icon-text-mic { 217 | font-size: 2.5em !important; 218 | } 219 | 220 | .icon-text-home { 221 | font-size: 3.5em !important; 222 | } 223 | 224 | .mic-button-div { 225 | position: fixed; 226 | left: 50%; 227 | top: 80% 228 | } 229 | 230 | /*############### Drop-down ############# */ 231 | .dropbtn { 232 | background-color: #ffffff; 233 | color: rgb(50, 71, 165); 234 | padding: 0px; 235 | font-size: 16px; 236 | border: none; 237 | } 238 | 239 | .dropdown { 240 | position: relative; 241 | display: inline-block; 242 | } 243 | 244 | .dropdown-content { 245 | display: none; 246 | position: absolute; 247 | background-color: #ffffff; 248 | min-width: 160px; 249 | box-shadow: 0px 8px 16px 0px rgba(0, 0, 0, 0.2); 250 | z-index: 1; 251 | } 252 | 253 | .dropdown-content a { 254 | color: black; 255 | padding: 12px 16px; 256 | text-decoration: none; 257 | display: block; 258 | } 259 | 260 | .dropdown-content a:hover { 261 | background-color: #ddd; 262 | } 263 | 264 | .dropdown:hover .dropdown-content { 265 | display: block; 266 | } 267 | 268 | .dropdown:hover .dropbtn { 269 | background-color: #3e8e41; 270 | } 271 | 272 | /* ############# Arrow ############## position: relative; position: absolute;*/ 273 | .load-more { 274 | position: fixed; 275 | cursor: pointer; 276 | width: 100px; 277 | height: 100px; 278 | margin: -0px 0 0 -0px; 279 | min-width: 10px; 280 | min-height: 10px; 281 | 282 | left: 90%; 283 | top: 45%; 284 | border-width: 2px; 285 | border-style: solid; 286 | border-color: transparent; 287 | border-bottom-color: #000; 288 | border-right-color: #000; 289 | border-radius: 0 0 5px 0; 290 | 291 | transform: translate(-0%, -0%) rotate(-45deg); 292 | } 293 | 294 | /* ######## Radio Buttons ############## */ 295 | .radio { 296 | background: #f6f7fd; 297 | padding: 4px; 298 | border-radius: 3px; 299 | box-shadow: inset 0 0 0 3px rgba(35, 33, 45, 0.3), 300 | 0 0 0 3px rgba(185, 185, 185, 0.3); 301 | position: relative; 302 | } 303 | 304 | .radio input { 305 | width: max-content; 306 | height: 100%; 307 | appearance: none; 308 | outline: none; 309 | cursor: pointer; 310 | border-radius: 2px; 311 | padding: 4px 8px; 312 | background: #454857; 313 | color: #bdbdbdbd; 314 | font-size: 0.8em; 315 | font-family: system-ui, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, 316 | "Helvetica Neue", Arial, "Noto Sans", sans-serif, "Apple Color Emoji", 317 | "Segoe UI Emoji", "Segoe UI Symbol", "Noto Color Emoji"; 318 | transition: all 100ms linear; 319 | } 320 | 321 | .radio input:checked { 322 | background-image: linear-gradient(180deg, #4e70ce, #5197d8); 323 | color: #fff; 324 | box-shadow: 0 1px 1px #0000002e; 325 | text-shadow: 0 1px 0px #79485f7a; 326 | } 327 | 328 | .radio input:before { 329 | content: attr(label); 330 | display: inline-block; 331 | text-align: center; 332 | width: 100%; 333 | } 334 | 335 | /* ############ Links and credits ####*/ 336 | 337 | .link-icon-div { 338 | position: fixed; 339 | left: 90.0%; 340 | top: 0.0%; 341 | vertical-align: middle; 342 | align-content: flex-start; 343 | } 344 | 345 | .credits-icon-div { 346 | position: fixed; 347 | left: 90.5%; 348 | top: 95%; 349 | font-size: x-small; 350 | } 351 | 352 | .svg-icon { 353 | padding-top: 1em; 354 | width: 50px; 355 | height: 50px; 356 | 357 | } 358 | 359 | 360 | /* ######## Switch ############## */ 361 | @media only screen and (max-width: 1200px) { 362 | .round-button { 363 | box-sizing: border-box; 364 | display: block; 365 | width: 2em; 366 | /* 80px */ 367 | height: 2em; 368 | left: -2.5%; 369 | padding-top: 0.3em; 370 | padding-left: 0px; 371 | line-height: 0px; 372 | border: 6px solid #fff; 373 | border-radius: 50%; 374 | color: #f5f5f5; 375 | text-align: center; 376 | text-decoration: none; 377 | background-color: #467387; 378 | font-size: 1em; 379 | font-weight: bold; 380 | transition: all 0.3s ease; 381 | } 382 | 383 | .container { 384 | display: block; 385 | position: absolute; 386 | left: 2%; 387 | top: 22%; 388 | transform: translate(-0%, -0%); 389 | height: 55%; 390 | width: 96%; 391 | max-width: 96%; 392 | background: #ffff; 393 | overflow: hidden; 394 | border-radius: 20px; 395 | box-shadow: 0 0 20px 8px #d0d0d0; 396 | } 397 | 398 | .icon-text { 399 | font-size: 0.8em !important; 400 | text-align: center; 401 | } 402 | 403 | .ipa-text-small { 404 | font-size: small; 405 | } 406 | 407 | .round-button-mic { 408 | box-sizing: border-box; 409 | display: block; 410 | width: 3.5em; 411 | /* 80px */ 412 | height: 3.5em; 413 | padding-top: 0.4em; 414 | left: 40%; 415 | line-height: 0px; 416 | border: 6px solid #fff; 417 | border-radius: 50%; 418 | color: #f5f5f5; 419 | text-align: center; 420 | text-decoration: none; 421 | background-color: #49d67d; 422 | font-size: 20px; 423 | font-weight: bold; 424 | transition: all 0.3s ease; 425 | } 426 | 427 | .mic-button-div { 428 | position: fixed; 429 | left: 40%; 430 | top: 80% 431 | } 432 | 433 | .link-icon-div { 434 | position: fixed; 435 | left: 89.0%; 436 | top: 0.0%; 437 | vertical-align: middle; 438 | } 439 | 440 | .credits-icon-div { 441 | position: fixed; 442 | left: 78.5%; 443 | top: 95%; 444 | font-size: x-small; 445 | } 446 | 447 | .svg-icon { 448 | padding-top: 1em; 449 | width: 40px; 450 | height: 40px; 451 | } 452 | 453 | 454 | .icon-text-home { 455 | font-size: 2.5em !important; 456 | } 457 | 458 | .accuracy-text { 459 | font-family: "Dank Mono", ui-monospace, monospace; 460 | background: linear-gradient(to right, 461 | rgb(54, 56, 80), 462 | rgb(21, 60, 87)); 463 | left: -5.0%; 464 | background-clip: text; 465 | -webkit-background-clip: text; 466 | -webkit-text-fill-color: transparent; 467 | text-align: center; 468 | font-size: 0.8em; 469 | } 470 | 471 | } -------------------------------------------------------------------------------- /static/javascript/callbacks.js: -------------------------------------------------------------------------------- 1 | 2 | 3 | // Audio context initialization 4 | let mediaRecorder, audioChunks, audioBlob, stream, audioRecorded; 5 | const ctx = new AudioContext(); 6 | let currentAudioForPlaying; 7 | let lettersOfWordAreCorrect = []; 8 | 9 | // UI-related variables 10 | const page_title = "AI Pronunciation Trainer"; 11 | const accuracy_colors = ["green", "orange", "red"]; 12 | let badScoreThreshold = 30; 13 | let mediumScoreThreshold = 70; 14 | let currentSample = 0; 15 | let currentScore = 0.; 16 | let sample_difficult = 0; 17 | let scoreMultiplier = 1; 18 | let playAnswerSounds = true; 19 | let isNativeSelectedForPlayback = true; 20 | let isRecording = false; 21 | let serverIsInitialized = false; 22 | let serverWorking = true; 23 | let languageFound = true; 24 | let currentSoundRecorded = false; 25 | let currentText, currentIpa, real_transcripts_ipa, matched_transcripts_ipa; 26 | let wordCategories; 27 | let startTime, endTime; 28 | 29 | // API related variables 30 | let AILanguage = "de"; // Standard is German 31 | 32 | 33 | let STScoreAPIKey = 'rll5QsTiv83nti99BW6uCmvs9BDVxSB39SVFceYb'; // Public Key. If, for some reason, you would like a private one, send-me a message and we can discuss some possibilities 34 | let apiMainPathSample = '';// 'http://127.0.0.1:3001';// 'https://a3hj0l2j2m.execute-api.eu-central-1.amazonaws.com/Prod'; 35 | let apiMainPathSTS = '';// 'https://wrg7ayuv7i.execute-api.eu-central-1.amazonaws.com/Prod'; 36 | 37 | 38 | // Variables to playback accuracy sounds 39 | let soundsPath = '../static';//'https://stscore-sounds-bucket.s3.eu-central-1.amazonaws.com'; 40 | let soundFileGood = null; 41 | let soundFileOkay = null; 42 | let soundFileBad = null; 43 | 44 | // Speech generation 45 | var synth = window.speechSynthesis; 46 | let voice_idx = 0; 47 | let voice_synth = null; 48 | 49 | //############################ UI general control functions ################### 50 | const unblockUI = () => { 51 | document.getElementById("recordAudio").classList.remove('disabled'); 52 | document.getElementById("playSampleAudio").classList.remove('disabled'); 53 | document.getElementById("buttonNext").onclick = () => getNextSample(); 54 | document.getElementById("nextButtonDiv").classList.remove('disabled'); 55 | document.getElementById("original_script").classList.remove('disabled'); 56 | document.getElementById("buttonNext").style["background-color"] = '#58636d'; 57 | 58 | if (currentSoundRecorded) 59 | document.getElementById("playRecordedAudio").classList.remove('disabled'); 60 | 61 | 62 | }; 63 | 64 | const blockUI = () => { 65 | 66 | document.getElementById("recordAudio").classList.add('disabled'); 67 | document.getElementById("playSampleAudio").classList.add('disabled'); 68 | document.getElementById("buttonNext").onclick = null; 69 | document.getElementById("original_script").classList.add('disabled'); 70 | document.getElementById("playRecordedAudio").classList.add('disabled'); 71 | 72 | document.getElementById("buttonNext").style["background-color"] = '#adadad'; 73 | 74 | 75 | }; 76 | 77 | const UIError = () => { 78 | blockUI(); 79 | document.getElementById("buttonNext").onclick = () => getNextSample(); //If error, user can only try to get a new sample 80 | document.getElementById("buttonNext").style["background-color"] = '#58636d'; 81 | 82 | document.getElementById("recorded_ipa_script").innerHTML = ""; 83 | document.getElementById("single_word_ipa_pair").innerHTML = "Error"; 84 | document.getElementById("ipa_script").innerHTML = "Error" 85 | 86 | document.getElementById("main_title").innerHTML = 'Server Error'; 87 | document.getElementById("original_script").innerHTML = 'Server error. Either the daily quota of the server is over or there was some internal error. You can try to generate a new sample in a few seconds. If the error persist, try comming back tomorrow or download the local version from Github :)'; 88 | }; 89 | 90 | const UINotSupported = () => { 91 | unblockUI(); 92 | 93 | document.getElementById("main_title").innerHTML = "Browser unsupported"; 94 | 95 | } 96 | 97 | const UIRecordingError = () => { 98 | unblockUI(); 99 | document.getElementById("main_title").innerHTML = "Recording error, please try again or restart page."; 100 | startMediaDevice(); 101 | } 102 | 103 | 104 | 105 | //################### Application state functions ####################### 106 | function updateScore(currentPronunciationScore) { 107 | 108 | if (isNaN(currentPronunciationScore)) 109 | return; 110 | currentScore += currentPronunciationScore * scoreMultiplier; 111 | currentScore = Math.round(currentScore); 112 | } 113 | 114 | const cacheSoundFiles = async () => { 115 | await fetch(soundsPath + '/ASR_good.wav').then(data => data.arrayBuffer()). 116 | then(arrayBuffer => ctx.decodeAudioData(arrayBuffer)). 117 | then(decodeAudioData => { 118 | soundFileGood = decodeAudioData; 119 | }); 120 | 121 | await fetch(soundsPath + '/ASR_okay.wav').then(data => data.arrayBuffer()). 122 | then(arrayBuffer => ctx.decodeAudioData(arrayBuffer)). 123 | then(decodeAudioData => { 124 | soundFileOkay = decodeAudioData; 125 | }); 126 | 127 | await fetch(soundsPath + '/ASR_bad.wav').then(data => data.arrayBuffer()). 128 | then(arrayBuffer => ctx.decodeAudioData(arrayBuffer)). 129 | then(decodeAudioData => { 130 | soundFileBad = decodeAudioData; 131 | }); 132 | } 133 | 134 | const getNextSample = async () => { 135 | 136 | 137 | 138 | blockUI(); 139 | 140 | if (!serverIsInitialized) 141 | await initializeServer(); 142 | 143 | if (!serverWorking) { 144 | UIError(); 145 | return; 146 | } 147 | 148 | if (soundFileBad == null) 149 | cacheSoundFiles(); 150 | 151 | 152 | 153 | updateScore(parseFloat(document.getElementById("pronunciation_accuracy").innerHTML)); 154 | 155 | document.getElementById("main_title").innerHTML = "Processing new sample..."; 156 | 157 | 158 | if (document.getElementById('lengthCat1').checked) { 159 | sample_difficult = 0; 160 | scoreMultiplier = 1.3; 161 | } 162 | else if (document.getElementById('lengthCat2').checked) { 163 | sample_difficult = 1; 164 | scoreMultiplier = 1; 165 | } 166 | else if (document.getElementById('lengthCat3').checked) { 167 | sample_difficult = 2; 168 | scoreMultiplier = 1.3; 169 | } 170 | else if (document.getElementById('lengthCat4').checked) { 171 | sample_difficult = 3; 172 | scoreMultiplier = 1.6; 173 | } 174 | 175 | try { 176 | await fetch(apiMainPathSample + '/getSample', { 177 | method: "post", 178 | body: JSON.stringify({ 179 | "category": sample_difficult.toString(), "language": AILanguage 180 | }), 181 | headers: { "X-Api-Key": STScoreAPIKey } 182 | }).then(res => res.json()). 183 | then(data => { 184 | 185 | 186 | 187 | let doc = document.getElementById("original_script"); 188 | currentText = data.real_transcript; 189 | doc.innerHTML = currentText; 190 | 191 | currentIpa = data.ipa_transcript 192 | 193 | let doc_ipa = document.getElementById("ipa_script"); 194 | doc_ipa.innerHTML = "/ " + currentIpa + " /"; 195 | 196 | document.getElementById("recorded_ipa_script").innerHTML = "" 197 | document.getElementById("pronunciation_accuracy").innerHTML = ""; 198 | document.getElementById("single_word_ipa_pair").innerHTML = "Reference | Spoken" 199 | document.getElementById("section_accuracy").innerHTML = "| Score: " + currentScore.toString() + " - (" + currentSample.toString() + ")"; 200 | currentSample += 1; 201 | 202 | document.getElementById("main_title").innerHTML = page_title; 203 | 204 | document.getElementById("translated_script").innerHTML = data.transcript_translation; 205 | 206 | currentSoundRecorded = false; 207 | unblockUI(); 208 | document.getElementById("playRecordedAudio").classList.add('disabled'); 209 | 210 | }) 211 | } 212 | catch 213 | { 214 | UIError(); 215 | } 216 | 217 | 218 | }; 219 | 220 | const updateRecordingState = async () => { 221 | if (isRecording) { 222 | stopRecording(); 223 | return 224 | } 225 | else { 226 | recordSample() 227 | return; 228 | } 229 | } 230 | 231 | const generateWordModal = (word_idx) => { 232 | 233 | document.getElementById("single_word_ipa_pair").innerHTML = wrapWordForPlayingLink(real_transcripts_ipa[word_idx], word_idx, false, "black") 234 | + ' | ' + wrapWordForPlayingLink(matched_transcripts_ipa[word_idx], word_idx, true, accuracy_colors[parseInt(wordCategories[word_idx])]) 235 | } 236 | 237 | const recordSample = async () => { 238 | 239 | document.getElementById("main_title").innerHTML = "Recording... click again when done speaking"; 240 | document.getElementById("recordIcon").innerHTML = 'pause_presentation'; 241 | blockUI(); 242 | document.getElementById("recordAudio").classList.remove('disabled'); 243 | audioChunks = []; 244 | isRecording = true; 245 | mediaRecorder.start(); 246 | 247 | } 248 | 249 | const changeLanguage = (language, generateNewSample = false) => { 250 | voices = synth.getVoices(); 251 | AILanguage = language; 252 | languageFound = false; 253 | let languageIdentifier, languageName; 254 | switch (language) { 255 | case 'de': 256 | 257 | document.getElementById("languageBox").innerHTML = "German"; 258 | languageIdentifier = 'de'; 259 | languageName = 'Anna'; 260 | break; 261 | 262 | case 'en': 263 | 264 | document.getElementById("languageBox").innerHTML = "English"; 265 | languageIdentifier = 'en'; 266 | languageName = 'Daniel'; 267 | break; 268 | }; 269 | 270 | for (idx = 0; idx < voices.length; idx++) { 271 | if (voices[idx].lang.slice(0, 2) == languageIdentifier && voices[idx].name == languageName) { 272 | voice_synth = voices[idx]; 273 | languageFound = true; 274 | break; 275 | } 276 | 277 | } 278 | // If specific voice not found, search anything with the same language 279 | if (!languageFound) { 280 | for (idx = 0; idx < voices.length; idx++) { 281 | if (voices[idx].lang.slice(0, 2) == languageIdentifier) { 282 | voice_synth = voices[idx]; 283 | languageFound = true; 284 | break; 285 | } 286 | } 287 | } 288 | if (generateNewSample) 289 | getNextSample(); 290 | } 291 | 292 | //################### Speech-To-Score function ######################## 293 | const mediaStreamConstraints = { 294 | audio: { 295 | channelCount: 1, 296 | sampleRate: 48000 297 | } 298 | } 299 | 300 | 301 | const startMediaDevice = () => { 302 | navigator.mediaDevices.getUserMedia(mediaStreamConstraints).then(_stream => { 303 | stream = _stream 304 | mediaRecorder = new MediaRecorder(stream); 305 | 306 | let currentSamples = 0 307 | mediaRecorder.ondataavailable = event => { 308 | 309 | currentSamples += event.data.length 310 | audioChunks.push(event.data); 311 | }; 312 | 313 | mediaRecorder.onstop = async () => { 314 | 315 | 316 | document.getElementById("recordIcon").innerHTML = 'mic'; 317 | blockUI(); 318 | 319 | 320 | audioBlob = new Blob(audioChunks, { type: 'audio/ogg;' }); 321 | 322 | let audioUrl = URL.createObjectURL(audioBlob); 323 | audioRecorded = new Audio(audioUrl); 324 | 325 | let audioBase64 = await convertBlobToBase64(audioBlob); 326 | 327 | let minimumAllowedLength = 6; 328 | if (audioBase64.length < minimumAllowedLength) { 329 | setTimeout(UIRecordingError, 50); // Make sure this function finished after get called again 330 | return; 331 | } 332 | 333 | try { 334 | // Get currentText from "original_script" div, in case user has change it 335 | let text = document.getElementById("original_script").innerHTML; 336 | // Remove html tags 337 | text = text.replace(/<[^>]*>?/gm, ''); 338 | //Remove spaces on the beginning and end 339 | text = text.trim(); 340 | // Remove double spaces 341 | text = text.replace(/\s\s+/g, ' '); 342 | currentText = [text]; 343 | 344 | await fetch(apiMainPathSTS + '/GetAccuracyFromRecordedAudio', { 345 | method: "post", 346 | body: JSON.stringify({ "title": currentText[0], "base64Audio": audioBase64, "language": AILanguage }), 347 | headers: { "X-Api-Key": STScoreAPIKey } 348 | 349 | }).then(res => res.json()). 350 | then(data => { 351 | 352 | if (playAnswerSounds) 353 | playSoundForAnswerAccuracy(parseFloat(data.pronunciation_accuracy)) 354 | 355 | document.getElementById("recorded_ipa_script").innerHTML = "/ " + data.ipa_transcript + " /"; 356 | document.getElementById("recordAudio").classList.add('disabled'); 357 | document.getElementById("main_title").innerHTML = page_title; 358 | document.getElementById("pronunciation_accuracy").innerHTML = data.pronunciation_accuracy + "%"; 359 | document.getElementById("ipa_script").innerHTML = data.real_transcripts_ipa 360 | 361 | lettersOfWordAreCorrect = data.is_letter_correct_all_words.split(" ") 362 | 363 | 364 | startTime = data.start_time; 365 | endTime = data.end_time; 366 | 367 | 368 | real_transcripts_ipa = data.real_transcripts_ipa.split(" ") 369 | matched_transcripts_ipa = data.matched_transcripts_ipa.split(" ") 370 | wordCategories = data.pair_accuracy_category.split(" ") 371 | let currentTextWords = currentText[0].split(" ") 372 | 373 | coloredWords = ""; 374 | for (let word_idx = 0; word_idx < currentTextWords.length; word_idx++) { 375 | 376 | wordTemp = ''; 377 | for (let letter_idx = 0; letter_idx < currentTextWords[word_idx].length; letter_idx++) { 378 | letter_is_correct = lettersOfWordAreCorrect[word_idx][letter_idx] == '1' 379 | if (letter_is_correct) 380 | color_letter = 'green' 381 | else 382 | color_letter = 'red' 383 | 384 | wordTemp += '' + currentTextWords[word_idx][letter_idx] + "" 385 | } 386 | currentTextWords[word_idx] 387 | coloredWords += " " + wrapWordForIndividualPlayback(wordTemp, word_idx) 388 | } 389 | 390 | 391 | 392 | document.getElementById("original_script").innerHTML = coloredWords 393 | 394 | currentSoundRecorded = true; 395 | unblockUI(); 396 | document.getElementById("playRecordedAudio").classList.remove('disabled'); 397 | 398 | }); 399 | } 400 | catch { 401 | UIError(); 402 | } 403 | }; 404 | 405 | }); 406 | }; 407 | startMediaDevice(); 408 | 409 | // ################### Audio playback ################## 410 | const playSoundForAnswerAccuracy = async (accuracy) => { 411 | 412 | currentAudioForPlaying = soundFileGood; 413 | if (accuracy < mediumScoreThreshold) { 414 | if (accuracy < badScoreThreshold) { 415 | currentAudioForPlaying = soundFileBad; 416 | } 417 | else { 418 | currentAudioForPlaying = soundFileOkay; 419 | } 420 | } 421 | playback(); 422 | 423 | } 424 | 425 | const playAudio = async () => { 426 | 427 | document.getElementById("main_title").innerHTML = "Generating sound..."; 428 | playWithMozillaApi(currentText[0]); 429 | document.getElementById("main_title").innerHTML = "Current Sound was played"; 430 | 431 | }; 432 | 433 | function playback() { 434 | const playSound = ctx.createBufferSource(); 435 | playSound.buffer = currentAudioForPlaying; 436 | playSound.connect(ctx.destination); 437 | playSound.start(ctx.currentTime) 438 | } 439 | 440 | 441 | const playRecording = async (start = null, end = null) => { 442 | blockUI(); 443 | 444 | try { 445 | if (start == null || end == null) { 446 | endTimeInMs = Math.round(audioRecorded.duration * 1000) 447 | audioRecorded.addEventListener("ended", function () { 448 | audioRecorded.currentTime = 0; 449 | unblockUI(); 450 | document.getElementById("main_title").innerHTML = "Recorded Sound was played"; 451 | }); 452 | await audioRecorded.play(); 453 | 454 | } 455 | else { 456 | audioRecorded.currentTime = start; 457 | audioRecorded.play(); 458 | durationInSeconds = end - start; 459 | endTimeInMs = Math.round(durationInSeconds * 1000); 460 | setTimeout(function () { 461 | unblockUI(); 462 | audioRecorded.pause(); 463 | audioRecorded.currentTime = 0; 464 | document.getElementById("main_title").innerHTML = "Recorded Sound was played"; 465 | }, endTimeInMs); 466 | 467 | } 468 | } 469 | catch { 470 | UINotSupported(); 471 | } 472 | }; 473 | 474 | const playNativeAndRecordedWord = async (word_idx) => { 475 | 476 | if (isNativeSelectedForPlayback) 477 | playCurrentWord(word_idx) 478 | else 479 | playRecordedWord(word_idx); 480 | 481 | isNativeSelectedForPlayback = !isNativeSelectedForPlayback; 482 | } 483 | 484 | const stopRecording = () => { 485 | isRecording = false 486 | mediaRecorder.stop() 487 | document.getElementById("main_title").innerHTML = "Processing audio..."; 488 | } 489 | 490 | 491 | const playCurrentWord = async (word_idx) => { 492 | 493 | document.getElementById("main_title").innerHTML = "Generating word..."; 494 | playWithMozillaApi(currentText[0].split(' ')[word_idx]); 495 | document.getElementById("main_title").innerHTML = "Word was played"; 496 | } 497 | 498 | // TODO: Check if fallback is correct 499 | const playWithMozillaApi = (text) => { 500 | 501 | if (languageFound) { 502 | blockUI(); 503 | if (voice_synth == null) 504 | changeLanguage(AILanguage); 505 | 506 | var utterThis = new SpeechSynthesisUtterance(text); 507 | utterThis.voice = voice_synth; 508 | utterThis.rate = 0.7; 509 | utterThis.onend = function (event) { 510 | unblockUI(); 511 | } 512 | synth.speak(utterThis); 513 | } 514 | else { 515 | UINotSupported(); 516 | } 517 | } 518 | 519 | const playRecordedWord = (word_idx) => { 520 | 521 | wordStartTime = parseFloat(startTime.split(' ')[word_idx]); 522 | wordEndTime = parseFloat(endTime.split(' ')[word_idx]); 523 | 524 | playRecording(wordStartTime, wordEndTime); 525 | 526 | } 527 | 528 | // ############# Utils ##################### 529 | const convertBlobToBase64 = async (blob) => { 530 | return await blobToBase64(blob); 531 | } 532 | 533 | const blobToBase64 = blob => new Promise((resolve, reject) => { 534 | const reader = new FileReader(); 535 | reader.readAsDataURL(blob); 536 | reader.onload = () => resolve(reader.result); 537 | reader.onerror = error => reject(error); 538 | }); 539 | 540 | const wrapWordForPlayingLink = (word, word_idx, isFromRecording, word_accuracy_color) => { 541 | if (isFromRecording) 542 | return '' + word + ' ' 543 | else 544 | return '' + word + ' ' 545 | } 546 | 547 | const wrapWordForIndividualPlayback = (word, word_idx) => { 548 | 549 | 550 | return '' + word + ' ' 551 | 552 | } 553 | 554 | // ########## Function to initialize server ############### 555 | // This is to try to avoid aws lambda cold start 556 | try { 557 | fetch(apiMainPathSTS + '/GetAccuracyFromRecordedAudio', { 558 | method: "post", 559 | body: JSON.stringify({ "title": '', "base64Audio": '', "language": AILanguage }), 560 | headers: { "X-Api-Key": STScoreAPIKey } 561 | 562 | }); 563 | } 564 | catch { } 565 | 566 | const initializeServer = async () => { 567 | 568 | valid_response = false; 569 | document.getElementById("main_title").innerHTML = 'Initializing server, this may take up to 2 minutes...'; 570 | let number_of_tries = 0; 571 | let maximum_number_of_tries = 4; 572 | 573 | while (!valid_response) { 574 | if (number_of_tries > maximum_number_of_tries) { 575 | serverWorking = false; 576 | break; 577 | } 578 | 579 | try { 580 | await fetch(apiMainPathSTS + '/GetAccuracyFromRecordedAudio', { 581 | method: "post", 582 | body: JSON.stringify({ "title": '', "base64Audio": '', "language": AILanguage }), 583 | headers: { "X-Api-Key": STScoreAPIKey } 584 | 585 | }).then( 586 | valid_response = true); 587 | serverIsInitialized = true; 588 | } 589 | catch 590 | { 591 | number_of_tries += 1; 592 | } 593 | } 594 | } 595 | 596 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU AFFERO GENERAL PUBLIC LICENSE 2 | Version 3, 19 November 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU Affero General Public License is a free, copyleft license for 11 | software and other kinds of works, specifically designed to ensure 12 | cooperation with the community in the case of network server software. 13 | 14 | The licenses for most software and other practical works are designed 15 | to take away your freedom to share and change the works. By contrast, 16 | our General Public Licenses are intended to guarantee your freedom to 17 | share and change all versions of a program--to make sure it remains free 18 | software for all its users. 19 | 20 | When we speak of free software, we are referring to freedom, not 21 | price. Our General Public Licenses are designed to make sure that you 22 | have the freedom to distribute copies of free software (and charge for 23 | them if you wish), that you receive source code or can get it if you 24 | want it, that you can change the software or use pieces of it in new 25 | free programs, and that you know you can do these things. 26 | 27 | Developers that use our General Public Licenses protect your rights 28 | with two steps: (1) assert copyright on the software, and (2) offer 29 | you this License which gives you legal permission to copy, distribute 30 | and/or modify the software. 31 | 32 | A secondary benefit of defending all users' freedom is that 33 | improvements made in alternate versions of the program, if they 34 | receive widespread use, become available for other developers to 35 | incorporate. Many developers of free software are heartened and 36 | encouraged by the resulting cooperation. However, in the case of 37 | software used on network servers, this result may fail to come about. 38 | The GNU General Public License permits making a modified version and 39 | letting the public access it on a server without ever releasing its 40 | source code to the public. 41 | 42 | The GNU Affero General Public License is designed specifically to 43 | ensure that, in such cases, the modified source code becomes available 44 | to the community. It requires the operator of a network server to 45 | provide the source code of the modified version running there to the 46 | users of that server. Therefore, public use of a modified version, on 47 | a publicly accessible server, gives the public access to the source 48 | code of the modified version. 49 | 50 | An older license, called the Affero General Public License and 51 | published by Affero, was designed to accomplish similar goals. This is 52 | a different license, not a version of the Affero GPL, but Affero has 53 | released a new version of the Affero GPL which permits relicensing under 54 | this license. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | TERMS AND CONDITIONS 60 | 61 | 0. Definitions. 62 | 63 | "This License" refers to version 3 of the GNU Affero General Public License. 64 | 65 | "Copyright" also means copyright-like laws that apply to other kinds of 66 | works, such as semiconductor masks. 67 | 68 | "The Program" refers to any copyrightable work licensed under this 69 | License. Each licensee is addressed as "you". "Licensees" and 70 | "recipients" may be individuals or organizations. 71 | 72 | To "modify" a work means to copy from or adapt all or part of the work 73 | in a fashion requiring copyright permission, other than the making of an 74 | exact copy. The resulting work is called a "modified version" of the 75 | earlier work or a work "based on" the earlier work. 76 | 77 | A "covered work" means either the unmodified Program or a work based 78 | on the Program. 79 | 80 | To "propagate" a work means to do anything with it that, without 81 | permission, would make you directly or secondarily liable for 82 | infringement under applicable copyright law, except executing it on a 83 | computer or modifying a private copy. Propagation includes copying, 84 | distribution (with or without modification), making available to the 85 | public, and in some countries other activities as well. 86 | 87 | To "convey" a work means any kind of propagation that enables other 88 | parties to make or receive copies. Mere interaction with a user through 89 | a computer network, with no transfer of a copy, is not conveying. 90 | 91 | An interactive user interface displays "Appropriate Legal Notices" 92 | to the extent that it includes a convenient and prominently visible 93 | feature that (1) displays an appropriate copyright notice, and (2) 94 | tells the user that there is no warranty for the work (except to the 95 | extent that warranties are provided), that licensees may convey the 96 | work under this License, and how to view a copy of this License. If 97 | the interface presents a list of user commands or options, such as a 98 | menu, a prominent item in the list meets this criterion. 99 | 100 | 1. Source Code. 101 | 102 | The "source code" for a work means the preferred form of the work 103 | for making modifications to it. "Object code" means any non-source 104 | form of a work. 105 | 106 | A "Standard Interface" means an interface that either is an official 107 | standard defined by a recognized standards body, or, in the case of 108 | interfaces specified for a particular programming language, one that 109 | is widely used among developers working in that language. 110 | 111 | The "System Libraries" of an executable work include anything, other 112 | than the work as a whole, that (a) is included in the normal form of 113 | packaging a Major Component, but which is not part of that Major 114 | Component, and (b) serves only to enable use of the work with that 115 | Major Component, or to implement a Standard Interface for which an 116 | implementation is available to the public in source code form. A 117 | "Major Component", in this context, means a major essential component 118 | (kernel, window system, and so on) of the specific operating system 119 | (if any) on which the executable work runs, or a compiler used to 120 | produce the work, or an object code interpreter used to run it. 121 | 122 | The "Corresponding Source" for a work in object code form means all 123 | the source code needed to generate, install, and (for an executable 124 | work) run the object code and to modify the work, including scripts to 125 | control those activities. However, it does not include the work's 126 | System Libraries, or general-purpose tools or generally available free 127 | programs which are used unmodified in performing those activities but 128 | which are not part of the work. For example, Corresponding Source 129 | includes interface definition files associated with source files for 130 | the work, and the source code for shared libraries and dynamically 131 | linked subprograms that the work is specifically designed to require, 132 | such as by intimate data communication or control flow between those 133 | subprograms and other parts of the work. 134 | 135 | The Corresponding Source need not include anything that users 136 | can regenerate automatically from other parts of the Corresponding 137 | Source. 138 | 139 | The Corresponding Source for a work in source code form is that 140 | same work. 141 | 142 | 2. Basic Permissions. 143 | 144 | All rights granted under this License are granted for the term of 145 | copyright on the Program, and are irrevocable provided the stated 146 | conditions are met. This License explicitly affirms your unlimited 147 | permission to run the unmodified Program. The output from running a 148 | covered work is covered by this License only if the output, given its 149 | content, constitutes a covered work. This License acknowledges your 150 | rights of fair use or other equivalent, as provided by copyright law. 151 | 152 | You may make, run and propagate covered works that you do not 153 | convey, without conditions so long as your license otherwise remains 154 | in force. You may convey covered works to others for the sole purpose 155 | of having them make modifications exclusively for you, or provide you 156 | with facilities for running those works, provided that you comply with 157 | the terms of this License in conveying all material for which you do 158 | not control copyright. Those thus making or running the covered works 159 | for you must do so exclusively on your behalf, under your direction 160 | and control, on terms that prohibit them from making any copies of 161 | your copyrighted material outside their relationship with you. 162 | 163 | Conveying under any other circumstances is permitted solely under 164 | the conditions stated below. Sublicensing is not allowed; section 10 165 | makes it unnecessary. 166 | 167 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 168 | 169 | No covered work shall be deemed part of an effective technological 170 | measure under any applicable law fulfilling obligations under article 171 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 172 | similar laws prohibiting or restricting circumvention of such 173 | measures. 174 | 175 | When you convey a covered work, you waive any legal power to forbid 176 | circumvention of technological measures to the extent such circumvention 177 | is effected by exercising rights under this License with respect to 178 | the covered work, and you disclaim any intention to limit operation or 179 | modification of the work as a means of enforcing, against the work's 180 | users, your or third parties' legal rights to forbid circumvention of 181 | technological measures. 182 | 183 | 4. Conveying Verbatim Copies. 184 | 185 | You may convey verbatim copies of the Program's source code as you 186 | receive it, in any medium, provided that you conspicuously and 187 | appropriately publish on each copy an appropriate copyright notice; 188 | keep intact all notices stating that this License and any 189 | non-permissive terms added in accord with section 7 apply to the code; 190 | keep intact all notices of the absence of any warranty; and give all 191 | recipients a copy of this License along with the Program. 192 | 193 | You may charge any price or no price for each copy that you convey, 194 | and you may offer support or warranty protection for a fee. 195 | 196 | 5. Conveying Modified Source Versions. 197 | 198 | You may convey a work based on the Program, or the modifications to 199 | produce it from the Program, in the form of source code under the 200 | terms of section 4, provided that you also meet all of these conditions: 201 | 202 | a) The work must carry prominent notices stating that you modified 203 | it, and giving a relevant date. 204 | 205 | b) The work must carry prominent notices stating that it is 206 | released under this License and any conditions added under section 207 | 7. This requirement modifies the requirement in section 4 to 208 | "keep intact all notices". 209 | 210 | c) You must license the entire work, as a whole, under this 211 | License to anyone who comes into possession of a copy. This 212 | License will therefore apply, along with any applicable section 7 213 | additional terms, to the whole of the work, and all its parts, 214 | regardless of how they are packaged. This License gives no 215 | permission to license the work in any other way, but it does not 216 | invalidate such permission if you have separately received it. 217 | 218 | d) If the work has interactive user interfaces, each must display 219 | Appropriate Legal Notices; however, if the Program has interactive 220 | interfaces that do not display Appropriate Legal Notices, your 221 | work need not make them do so. 222 | 223 | A compilation of a covered work with other separate and independent 224 | works, which are not by their nature extensions of the covered work, 225 | and which are not combined with it such as to form a larger program, 226 | in or on a volume of a storage or distribution medium, is called an 227 | "aggregate" if the compilation and its resulting copyright are not 228 | used to limit the access or legal rights of the compilation's users 229 | beyond what the individual works permit. Inclusion of a covered work 230 | in an aggregate does not cause this License to apply to the other 231 | parts of the aggregate. 232 | 233 | 6. Conveying Non-Source Forms. 234 | 235 | You may convey a covered work in object code form under the terms 236 | of sections 4 and 5, provided that you also convey the 237 | machine-readable Corresponding Source under the terms of this License, 238 | in one of these ways: 239 | 240 | a) Convey the object code in, or embodied in, a physical product 241 | (including a physical distribution medium), accompanied by the 242 | Corresponding Source fixed on a durable physical medium 243 | customarily used for software interchange. 244 | 245 | b) Convey the object code in, or embodied in, a physical product 246 | (including a physical distribution medium), accompanied by a 247 | written offer, valid for at least three years and valid for as 248 | long as you offer spare parts or customer support for that product 249 | model, to give anyone who possesses the object code either (1) a 250 | copy of the Corresponding Source for all the software in the 251 | product that is covered by this License, on a durable physical 252 | medium customarily used for software interchange, for a price no 253 | more than your reasonable cost of physically performing this 254 | conveying of source, or (2) access to copy the 255 | Corresponding Source from a network server at no charge. 256 | 257 | c) Convey individual copies of the object code with a copy of the 258 | written offer to provide the Corresponding Source. This 259 | alternative is allowed only occasionally and noncommercially, and 260 | only if you received the object code with such an offer, in accord 261 | with subsection 6b. 262 | 263 | d) Convey the object code by offering access from a designated 264 | place (gratis or for a charge), and offer equivalent access to the 265 | Corresponding Source in the same way through the same place at no 266 | further charge. You need not require recipients to copy the 267 | Corresponding Source along with the object code. If the place to 268 | copy the object code is a network server, the Corresponding Source 269 | may be on a different server (operated by you or a third party) 270 | that supports equivalent copying facilities, provided you maintain 271 | clear directions next to the object code saying where to find the 272 | Corresponding Source. Regardless of what server hosts the 273 | Corresponding Source, you remain obligated to ensure that it is 274 | available for as long as needed to satisfy these requirements. 275 | 276 | e) Convey the object code using peer-to-peer transmission, provided 277 | you inform other peers where the object code and Corresponding 278 | Source of the work are being offered to the general public at no 279 | charge under subsection 6d. 280 | 281 | A separable portion of the object code, whose source code is excluded 282 | from the Corresponding Source as a System Library, need not be 283 | included in conveying the object code work. 284 | 285 | A "User Product" is either (1) a "consumer product", which means any 286 | tangible personal property which is normally used for personal, family, 287 | or household purposes, or (2) anything designed or sold for incorporation 288 | into a dwelling. In determining whether a product is a consumer product, 289 | doubtful cases shall be resolved in favor of coverage. For a particular 290 | product received by a particular user, "normally used" refers to a 291 | typical or common use of that class of product, regardless of the status 292 | of the particular user or of the way in which the particular user 293 | actually uses, or expects or is expected to use, the product. A product 294 | is a consumer product regardless of whether the product has substantial 295 | commercial, industrial or non-consumer uses, unless such uses represent 296 | the only significant mode of use of the product. 297 | 298 | "Installation Information" for a User Product means any methods, 299 | procedures, authorization keys, or other information required to install 300 | and execute modified versions of a covered work in that User Product from 301 | a modified version of its Corresponding Source. The information must 302 | suffice to ensure that the continued functioning of the modified object 303 | code is in no case prevented or interfered with solely because 304 | modification has been made. 305 | 306 | If you convey an object code work under this section in, or with, or 307 | specifically for use in, a User Product, and the conveying occurs as 308 | part of a transaction in which the right of possession and use of the 309 | User Product is transferred to the recipient in perpetuity or for a 310 | fixed term (regardless of how the transaction is characterized), the 311 | Corresponding Source conveyed under this section must be accompanied 312 | by the Installation Information. But this requirement does not apply 313 | if neither you nor any third party retains the ability to install 314 | modified object code on the User Product (for example, the work has 315 | been installed in ROM). 316 | 317 | The requirement to provide Installation Information does not include a 318 | requirement to continue to provide support service, warranty, or updates 319 | for a work that has been modified or installed by the recipient, or for 320 | the User Product in which it has been modified or installed. Access to a 321 | network may be denied when the modification itself materially and 322 | adversely affects the operation of the network or violates the rules and 323 | protocols for communication across the network. 324 | 325 | Corresponding Source conveyed, and Installation Information provided, 326 | in accord with this section must be in a format that is publicly 327 | documented (and with an implementation available to the public in 328 | source code form), and must require no special password or key for 329 | unpacking, reading or copying. 330 | 331 | 7. Additional Terms. 332 | 333 | "Additional permissions" are terms that supplement the terms of this 334 | License by making exceptions from one or more of its conditions. 335 | Additional permissions that are applicable to the entire Program shall 336 | be treated as though they were included in this License, to the extent 337 | that they are valid under applicable law. If additional permissions 338 | apply only to part of the Program, that part may be used separately 339 | under those permissions, but the entire Program remains governed by 340 | this License without regard to the additional permissions. 341 | 342 | When you convey a copy of a covered work, you may at your option 343 | remove any additional permissions from that copy, or from any part of 344 | it. (Additional permissions may be written to require their own 345 | removal in certain cases when you modify the work.) You may place 346 | additional permissions on material, added by you to a covered work, 347 | for which you have or can give appropriate copyright permission. 348 | 349 | Notwithstanding any other provision of this License, for material you 350 | add to a covered work, you may (if authorized by the copyright holders of 351 | that material) supplement the terms of this License with terms: 352 | 353 | a) Disclaiming warranty or limiting liability differently from the 354 | terms of sections 15 and 16 of this License; or 355 | 356 | b) Requiring preservation of specified reasonable legal notices or 357 | author attributions in that material or in the Appropriate Legal 358 | Notices displayed by works containing it; or 359 | 360 | c) Prohibiting misrepresentation of the origin of that material, or 361 | requiring that modified versions of such material be marked in 362 | reasonable ways as different from the original version; or 363 | 364 | d) Limiting the use for publicity purposes of names of licensors or 365 | authors of the material; or 366 | 367 | e) Declining to grant rights under trademark law for use of some 368 | trade names, trademarks, or service marks; or 369 | 370 | f) Requiring indemnification of licensors and authors of that 371 | material by anyone who conveys the material (or modified versions of 372 | it) with contractual assumptions of liability to the recipient, for 373 | any liability that these contractual assumptions directly impose on 374 | those licensors and authors. 375 | 376 | All other non-permissive additional terms are considered "further 377 | restrictions" within the meaning of section 10. If the Program as you 378 | received it, or any part of it, contains a notice stating that it is 379 | governed by this License along with a term that is a further 380 | restriction, you may remove that term. If a license document contains 381 | a further restriction but permits relicensing or conveying under this 382 | License, you may add to a covered work material governed by the terms 383 | of that license document, provided that the further restriction does 384 | not survive such relicensing or conveying. 385 | 386 | If you add terms to a covered work in accord with this section, you 387 | must place, in the relevant source files, a statement of the 388 | additional terms that apply to those files, or a notice indicating 389 | where to find the applicable terms. 390 | 391 | Additional terms, permissive or non-permissive, may be stated in the 392 | form of a separately written license, or stated as exceptions; 393 | the above requirements apply either way. 394 | 395 | 8. Termination. 396 | 397 | You may not propagate or modify a covered work except as expressly 398 | provided under this License. Any attempt otherwise to propagate or 399 | modify it is void, and will automatically terminate your rights under 400 | this License (including any patent licenses granted under the third 401 | paragraph of section 11). 402 | 403 | However, if you cease all violation of this License, then your 404 | license from a particular copyright holder is reinstated (a) 405 | provisionally, unless and until the copyright holder explicitly and 406 | finally terminates your license, and (b) permanently, if the copyright 407 | holder fails to notify you of the violation by some reasonable means 408 | prior to 60 days after the cessation. 409 | 410 | Moreover, your license from a particular copyright holder is 411 | reinstated permanently if the copyright holder notifies you of the 412 | violation by some reasonable means, this is the first time you have 413 | received notice of violation of this License (for any work) from that 414 | copyright holder, and you cure the violation prior to 30 days after 415 | your receipt of the notice. 416 | 417 | Termination of your rights under this section does not terminate the 418 | licenses of parties who have received copies or rights from you under 419 | this License. If your rights have been terminated and not permanently 420 | reinstated, you do not qualify to receive new licenses for the same 421 | material under section 10. 422 | 423 | 9. Acceptance Not Required for Having Copies. 424 | 425 | You are not required to accept this License in order to receive or 426 | run a copy of the Program. Ancillary propagation of a covered work 427 | occurring solely as a consequence of using peer-to-peer transmission 428 | to receive a copy likewise does not require acceptance. However, 429 | nothing other than this License grants you permission to propagate or 430 | modify any covered work. These actions infringe copyright if you do 431 | not accept this License. Therefore, by modifying or propagating a 432 | covered work, you indicate your acceptance of this License to do so. 433 | 434 | 10. Automatic Licensing of Downstream Recipients. 435 | 436 | Each time you convey a covered work, the recipient automatically 437 | receives a license from the original licensors, to run, modify and 438 | propagate that work, subject to this License. You are not responsible 439 | for enforcing compliance by third parties with this License. 440 | 441 | An "entity transaction" is a transaction transferring control of an 442 | organization, or substantially all assets of one, or subdividing an 443 | organization, or merging organizations. If propagation of a covered 444 | work results from an entity transaction, each party to that 445 | transaction who receives a copy of the work also receives whatever 446 | licenses to the work the party's predecessor in interest had or could 447 | give under the previous paragraph, plus a right to possession of the 448 | Corresponding Source of the work from the predecessor in interest, if 449 | the predecessor has it or can get it with reasonable efforts. 450 | 451 | You may not impose any further restrictions on the exercise of the 452 | rights granted or affirmed under this License. For example, you may 453 | not impose a license fee, royalty, or other charge for exercise of 454 | rights granted under this License, and you may not initiate litigation 455 | (including a cross-claim or counterclaim in a lawsuit) alleging that 456 | any patent claim is infringed by making, using, selling, offering for 457 | sale, or importing the Program or any portion of it. 458 | 459 | 11. Patents. 460 | 461 | A "contributor" is a copyright holder who authorizes use under this 462 | License of the Program or a work on which the Program is based. The 463 | work thus licensed is called the contributor's "contributor version". 464 | 465 | A contributor's "essential patent claims" are all patent claims 466 | owned or controlled by the contributor, whether already acquired or 467 | hereafter acquired, that would be infringed by some manner, permitted 468 | by this License, of making, using, or selling its contributor version, 469 | but do not include claims that would be infringed only as a 470 | consequence of further modification of the contributor version. For 471 | purposes of this definition, "control" includes the right to grant 472 | patent sublicenses in a manner consistent with the requirements of 473 | this License. 474 | 475 | Each contributor grants you a non-exclusive, worldwide, royalty-free 476 | patent license under the contributor's essential patent claims, to 477 | make, use, sell, offer for sale, import and otherwise run, modify and 478 | propagate the contents of its contributor version. 479 | 480 | In the following three paragraphs, a "patent license" is any express 481 | agreement or commitment, however denominated, not to enforce a patent 482 | (such as an express permission to practice a patent or covenant not to 483 | sue for patent infringement). To "grant" such a patent license to a 484 | party means to make such an agreement or commitment not to enforce a 485 | patent against the party. 486 | 487 | If you convey a covered work, knowingly relying on a patent license, 488 | and the Corresponding Source of the work is not available for anyone 489 | to copy, free of charge and under the terms of this License, through a 490 | publicly available network server or other readily accessible means, 491 | then you must either (1) cause the Corresponding Source to be so 492 | available, or (2) arrange to deprive yourself of the benefit of the 493 | patent license for this particular work, or (3) arrange, in a manner 494 | consistent with the requirements of this License, to extend the patent 495 | license to downstream recipients. "Knowingly relying" means you have 496 | actual knowledge that, but for the patent license, your conveying the 497 | covered work in a country, or your recipient's use of the covered work 498 | in a country, would infringe one or more identifiable patents in that 499 | country that you have reason to believe are valid. 500 | 501 | If, pursuant to or in connection with a single transaction or 502 | arrangement, you convey, or propagate by procuring conveyance of, a 503 | covered work, and grant a patent license to some of the parties 504 | receiving the covered work authorizing them to use, propagate, modify 505 | or convey a specific copy of the covered work, then the patent license 506 | you grant is automatically extended to all recipients of the covered 507 | work and works based on it. 508 | 509 | A patent license is "discriminatory" if it does not include within 510 | the scope of its coverage, prohibits the exercise of, or is 511 | conditioned on the non-exercise of one or more of the rights that are 512 | specifically granted under this License. You may not convey a covered 513 | work if you are a party to an arrangement with a third party that is 514 | in the business of distributing software, under which you make payment 515 | to the third party based on the extent of your activity of conveying 516 | the work, and under which the third party grants, to any of the 517 | parties who would receive the covered work from you, a discriminatory 518 | patent license (a) in connection with copies of the covered work 519 | conveyed by you (or copies made from those copies), or (b) primarily 520 | for and in connection with specific products or compilations that 521 | contain the covered work, unless you entered into that arrangement, 522 | or that patent license was granted, prior to 28 March 2007. 523 | 524 | Nothing in this License shall be construed as excluding or limiting 525 | any implied license or other defenses to infringement that may 526 | otherwise be available to you under applicable patent law. 527 | 528 | 12. No Surrender of Others' Freedom. 529 | 530 | If conditions are imposed on you (whether by court order, agreement or 531 | otherwise) that contradict the conditions of this License, they do not 532 | excuse you from the conditions of this License. If you cannot convey a 533 | covered work so as to satisfy simultaneously your obligations under this 534 | License and any other pertinent obligations, then as a consequence you may 535 | not convey it at all. For example, if you agree to terms that obligate you 536 | to collect a royalty for further conveying from those to whom you convey 537 | the Program, the only way you could satisfy both those terms and this 538 | License would be to refrain entirely from conveying the Program. 539 | 540 | 13. Remote Network Interaction; Use with the GNU General Public License. 541 | 542 | Notwithstanding any other provision of this License, if you modify the 543 | Program, your modified version must prominently offer all users 544 | interacting with it remotely through a computer network (if your version 545 | supports such interaction) an opportunity to receive the Corresponding 546 | Source of your version by providing access to the Corresponding Source 547 | from a network server at no charge, through some standard or customary 548 | means of facilitating copying of software. This Corresponding Source 549 | shall include the Corresponding Source for any work covered by version 3 550 | of the GNU General Public License that is incorporated pursuant to the 551 | following paragraph. 552 | 553 | Notwithstanding any other provision of this License, you have 554 | permission to link or combine any covered work with a work licensed 555 | under version 3 of the GNU General Public License into a single 556 | combined work, and to convey the resulting work. The terms of this 557 | License will continue to apply to the part which is the covered work, 558 | but the work with which it is combined will remain governed by version 559 | 3 of the GNU General Public License. 560 | 561 | 14. Revised Versions of this License. 562 | 563 | The Free Software Foundation may publish revised and/or new versions of 564 | the GNU Affero General Public License from time to time. Such new versions 565 | will be similar in spirit to the present version, but may differ in detail to 566 | address new problems or concerns. 567 | 568 | Each version is given a distinguishing version number. If the 569 | Program specifies that a certain numbered version of the GNU Affero General 570 | Public License "or any later version" applies to it, you have the 571 | option of following the terms and conditions either of that numbered 572 | version or of any later version published by the Free Software 573 | Foundation. If the Program does not specify a version number of the 574 | GNU Affero General Public License, you may choose any version ever published 575 | by the Free Software Foundation. 576 | 577 | If the Program specifies that a proxy can decide which future 578 | versions of the GNU Affero General Public License can be used, that proxy's 579 | public statement of acceptance of a version permanently authorizes you 580 | to choose that version for the Program. 581 | 582 | Later license versions may give you additional or different 583 | permissions. However, no additional obligations are imposed on any 584 | author or copyright holder as a result of your choosing to follow a 585 | later version. 586 | 587 | 15. Disclaimer of Warranty. 588 | 589 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 590 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 591 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 592 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 593 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 594 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 595 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 596 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 597 | 598 | 16. Limitation of Liability. 599 | 600 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 601 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 602 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 603 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 604 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 605 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 606 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 607 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 608 | SUCH DAMAGES. 609 | 610 | 17. Interpretation of Sections 15 and 16. 611 | 612 | If the disclaimer of warranty and limitation of liability provided 613 | above cannot be given local legal effect according to their terms, 614 | reviewing courts shall apply local law that most closely approximates 615 | an absolute waiver of all civil liability in connection with the 616 | Program, unless a warranty or assumption of liability accompanies a 617 | copy of the Program in return for a fee. 618 | 619 | END OF TERMS AND CONDITIONS 620 | 621 | How to Apply These Terms to Your New Programs 622 | 623 | If you develop a new program, and you want it to be of the greatest 624 | possible use to the public, the best way to achieve this is to make it 625 | free software which everyone can redistribute and change under these terms. 626 | 627 | To do so, attach the following notices to the program. It is safest 628 | to attach them to the start of each source file to most effectively 629 | state the exclusion of warranty; and each file should have at least 630 | the "copyright" line and a pointer to where the full notice is found. 631 | 632 | 633 | Copyright (C) 634 | 635 | This program is free software: you can redistribute it and/or modify 636 | it under the terms of the GNU Affero General Public License as published 637 | by the Free Software Foundation, either version 3 of the License, or 638 | (at your option) any later version. 639 | 640 | This program is distributed in the hope that it will be useful, 641 | but WITHOUT ANY WARRANTY; without even the implied warranty of 642 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 643 | GNU Affero General Public License for more details. 644 | 645 | You should have received a copy of the GNU Affero General Public License 646 | along with this program. If not, see . 647 | 648 | Also add information on how to contact you by electronic and paper mail. 649 | 650 | If your software can interact with users remotely through a computer 651 | network, you should also make sure that it provides a way for users to 652 | get its source. For example, if your program is a web application, its 653 | interface could display a "Source" link that leads users to an archive 654 | of the code. There are many ways you could offer source, and different 655 | solutions will be better for different programs; see section 13 for the 656 | specific requirements. 657 | 658 | You should also get your employer (if you work as a programmer) or school, 659 | if any, to sign a "copyright disclaimer" for the program, if necessary. 660 | For more information on this, and how to apply and follow the GNU AGPL, see 661 | . 662 | --------------------------------------------------------------------------------