├── .gitignore ├── .gitattributes ├── README.md ├── mce_utils.py ├── mce2018_dae_tst.py └── LICENSE /.gitignore: -------------------------------------------------------------------------------- 1 | data/*.csv 2 | *.pyc 3 | -------------------------------------------------------------------------------- /.gitattributes: -------------------------------------------------------------------------------- 1 | # Auto detect text files and perform LF normalization 2 | * text=auto 3 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # A DENOISING AUTOENCODER FOR SPEAKER IDENTIFICATION 2 | 3 | This is a Python implementation of the Denoising Autoencoder approach that we proposed for the first Multi-target speaker detection and identification Challenge Evaluation (MCE 2018, [http://www.mce2018.org](http://www.mce2018.org) ). 4 | 5 | The basic idea is to train a Denoising Autoencoder to map each individual input ivector to the mean of all ivectors from that speaker. The aim of this DAE is to compensate for inter-session variability and increase the discriminative power of the ivectors. 6 | 7 | You can find our system description for the MCE 2018 challenge [here](http://mce.csail.mit.edu/pdfs/BiometricVox_description.pdf). 8 | 9 | ## ABOUT THE MCE 2018 CHALLENGE 10 | 11 | The task for the MCE 2018 Evaluation was to detect if a given speech segment belongs to any of the speakers in a blacklist. The challenge is divided into two related subtasks: Top-S detection, i.e. detecting if the segment belongs to any of the blacklist speakers; and Top-1 detection, i.e. detecting which specific blacklist speaker (if any) is speaking in the segment. The data was generated from real call center user-agent telephone conversations. Instead of raw audio data, organizers processed the original data and provided 600-dimensional ivectors. This way, no special signal processing knowledge was needed to enter the evaluation. More details can be found on the [evaluation plan](https://arxiv.org/abs/1807.06663). 12 | 13 | ## DATASET 14 | 15 | The dataset can be found at: 16 | 17 | [https://www.kaggle.com/kagglesre/blacklist-speakers-dataset](https://www.kaggle.com/kagglesre/blacklist-speakers-dataset) 18 | 19 | After download, extract the files to _data_ folder. 20 | 21 | ## SYSTEM TRAINING 22 | 23 | Our training script shows how a very simple DAE can bring a very nice improvement over the [baseline](https://github.com/swshon/multi-speakerID). If you run 24 | 25 | ``` 26 | python mce2018_dae_tst.py 27 | ``` 28 | 29 | you should get results like these: 30 | 31 | ``` 32 | Dev set score using train set : 33 | Top S detector EER is 2.40% 34 | Top 1 detector EER is 9.50% (Total confusion error is 343) 35 | 36 | Test set score using train set: 37 | Top S detector EER is 6.83% 38 | Top 1 detector EER is 12.42% (Total confusion error is 411) 39 | 40 | Test set score using train + dev set: 41 | Top S detector EER is 5.69% 42 | Top 1 detector EER is 8.90% (Total confusion error is 257) 43 | ``` 44 | 45 | Note that these results do not match our official submission to the challenge, were we obtained Top-S EER: 4.33%, Top-1 EER: 6.11%, since our final system was a bit more complex including Probabilistic Linear Discriminant Analysis (PLDA) scoring and Symmetric Normalization (S-Norm). 46 | 47 | ## Requirements 48 | 49 | * Numpy 50 | * Scikit-learn 51 | * Pandas 52 | * Keras 53 | 54 | The code should run on both Python 2 and 3. 55 | -------------------------------------------------------------------------------- /mce_utils.py: -------------------------------------------------------------------------------- 1 | """ 2 | Utility functions adapted from: https://github.com/swshon/multi-speakerID. 3 | 4 | Copyright 2018 Roberto Font 5 | Biometric Vox S.L. 6 | 7 | Licensed under the Apache License, Version 2.0 (the "License"); 8 | you may not use this file except in compliance with the License. 9 | You may obtain a copy of the License at 10 | 11 | http://www.apache.org/licenses/LICENSE-2.0 12 | 13 | Unless required by applicable law or agreed to in writing, software 14 | distributed under the License is distributed on an "AS IS" BASIS, 15 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 16 | See the License for the specific language governing permissions and 17 | limitations under the License. 18 | """ 19 | 20 | from __future__ import print_function 21 | import numpy as np 22 | from sklearn.metrics import roc_curve 23 | 24 | def load_ivector(filename): 25 | utt = np.loadtxt(filename,dtype='str',delimiter=',',skiprows=1,usecols=[0]) 26 | ivector = np.loadtxt(filename,dtype='float32',delimiter=',',skiprows=1,usecols=range(1,601)) 27 | spk_id = [] 28 | for iter in range(len(utt)): 29 | spk_id = np.append(spk_id,utt[iter].split('_')[0]) 30 | 31 | return spk_id, utt, ivector 32 | 33 | def length_norm(mat): 34 | # length normalization (l2 norm) 35 | # input: mat = [utterances X vector dimension] ex) (float) 8631 X 600 36 | 37 | norm_mat = [] 38 | for line in mat: 39 | temp = line/np.math.sqrt(sum(np.power(line,2))) 40 | norm_mat.append(temp) 41 | norm_mat = np.array(norm_mat) 42 | return norm_mat 43 | 44 | def make_spkvec(mat, spk_label): 45 | # calculating speaker mean vector 46 | # input: mat = [utterances X vector dimension] ex) (float) 8631 X 600 47 | # spk_label = string vector ex) ['abce','cdgd'] 48 | 49 | # for iter in range(len(spk_label)): 50 | # spk_label[iter] = spk_label[iter].split('_')[0] 51 | 52 | spk_label, spk_index = np.unique(spk_label,return_inverse=True) 53 | spk_mean=[] 54 | mat = np.array(mat) 55 | 56 | # calculating speaker mean i-vector 57 | for i, spk in enumerate(spk_label): 58 | spk_mean.append(np.mean(mat[np.nonzero(spk_index==i)],axis=0)) 59 | spk_mean = length_norm(spk_mean) 60 | return spk_mean, spk_label 61 | 62 | def calculate_EER(trials, scores): 63 | # calculating EER of Top-S detector 64 | # input: trials = boolean(or int) vector, 1: postive(blacklist) 0: negative(background) 65 | # scores = float vector 66 | 67 | # Calculating EER 68 | fpr,tpr,threshold = roc_curve(trials,scores,pos_label=1) 69 | fnr = 1-tpr 70 | EER_threshold = threshold[np.argmin(abs(fnr-fpr))] 71 | 72 | # print EER_threshold 73 | EER_fpr = fpr[np.argmin(np.absolute((fnr-fpr)))] 74 | EER_fnr = fnr[np.argmin(np.absolute((fnr-fpr)))] 75 | EER = 0.5 * (EER_fpr+EER_fnr) 76 | 77 | print("Top S detector EER is %0.2f%%"% (EER*100)) 78 | return EER 79 | 80 | def get_trials_label_with_confusion(identified_label, groundtruth_label,dict4spk,is_trial ): 81 | # determine if the test utterance would make confusion error 82 | # input: identified_label = string vector, identified result of test utterance among multi-target from the detection system 83 | # groundtruth_label = string vector, ground truth speaker labels of test utterances 84 | # dict4spk = dictionary, convert label to target set, ex) train2dev convert train id to dev id 85 | 86 | trials = np.zeros(len(identified_label)) 87 | for iter in range(0,len(groundtruth_label)): 88 | enroll = identified_label[iter].split('_')[0] 89 | test = groundtruth_label[iter].split('_')[0] 90 | if is_trial[iter]: 91 | if enroll == dict4spk[test]: 92 | trials[iter]=1 # for Target trial (blacklist speaker) 93 | else: 94 | trials[iter]=-1 # for Target trial (backlist speaker), but fail on blacklist classifier 95 | 96 | else : 97 | trials[iter]=0 # for non-target (non-blacklist speaker) 98 | return trials 99 | 100 | 101 | def calculate_EER_with_confusion(scores,trials): 102 | # calculating EER of Top-1 detector 103 | # input: trials = boolean(or int) vector, 1: postive(blacklist) 0: negative(background) -1: confusion(blacklist) 104 | # scores = float vector 105 | 106 | # exclude confusion error (trials==-1) 107 | scores_wo_confusion = scores[np.nonzero(trials!=-1)[0]] 108 | trials_wo_confusion = trials[np.nonzero(trials!=-1)[0]] 109 | 110 | # dev_trials contain labels of target. (target=1, non-target=0) 111 | fpr,tpr,threshold = roc_curve(trials_wo_confusion,scores_wo_confusion,pos_label=1, drop_intermediate=False) 112 | fnr = 1-tpr 113 | EER_threshold = threshold[np.argmin(abs(fnr-fpr))] 114 | 115 | # EER withouth confusion error 116 | EER = fpr[np.argmin(np.absolute((fnr-fpr)))] 117 | 118 | # Add confusion error to false negative rate(Miss rate) 119 | total_negative = len(np.nonzero(np.array(trials_wo_confusion)==0)[0]) 120 | total_positive = len(np.nonzero(np.array(trials_wo_confusion)==1)[0]) 121 | fp= fpr*np.float(total_negative) 122 | fn= fnr*np.float(total_positive) 123 | fn += len(np.nonzero(trials==-1)[0]) 124 | total_positive += len(np.nonzero(trials==-1)[0]) 125 | fpr= fp/total_negative 126 | fnr= fn/total_positive 127 | 128 | # EER with confusion Error 129 | EER_threshold = threshold[np.argmin(abs(fnr-fpr))] 130 | EER_fpr = fpr[np.argmin(np.absolute((fnr-fpr)))] 131 | EER_fnr = fnr[np.argmin(np.absolute((fnr-fpr)))] 132 | EER = 0.5 * (EER_fpr+EER_fnr) 133 | 134 | print("Top 1 detector EER is %0.2f%% (Total confusion error is %d)"% ((EER*100), len(np.nonzero(trials==-1)[0]))) 135 | return EER 136 | -------------------------------------------------------------------------------- /mce2018_dae_tst.py: -------------------------------------------------------------------------------- 1 | """ 2 | This is a Python implementation of the Denoising Autoencoder approach that we proposed for 3 | the first Multi-target speaker detection and identification Challenge Evaluation (MCE 2018, http://www.mce2018.org) 4 | 5 | The basic idea is to train a Denoising Autoencoder to map each individual input ivector 6 | to the mean of all ivectors from that speaker. 7 | The aim of this DAE is to compensate for inter-session variability and increase the discriminative power of the ivectors. 8 | 9 | You can find our system description for the MCE 2018 challenge here: http://mce.csail.mit.edu/pdfs/BiometricVox_description.pdf 10 | 11 | Part of the code has been adapted from the baseline system at: https://github.com/swshon/multi-speakerID. 12 | 13 | Copyright 2018 Roberto Font 14 | Biometric Vox S.L. 15 | 16 | Licensed under the Apache License, Version 2.0 (the "License"); 17 | you may not use this file except in compliance with the License. 18 | You may obtain a copy of the License at 19 | 20 | http://www.apache.org/licenses/LICENSE-2.0 21 | 22 | Unless required by applicable law or agreed to in writing, software 23 | distributed under the License is distributed on an "AS IS" BASIS, 24 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 25 | See the License for the specific language governing permissions and 26 | limitations under the License. 27 | """ 28 | 29 | from __future__ import print_function 30 | import numpy as np 31 | 32 | from mce_utils import load_ivector, length_norm, make_spkvec, calculate_EER, get_trials_label_with_confusion, calculate_EER_with_confusion 33 | 34 | import pandas as pd 35 | from keras.models import Model 36 | from keras.layers import Input, Dense, Activation 37 | from keras import metrics 38 | from keras import optimizers 39 | 40 | # Neural network definition: a single hidden layer with 'tanh' activation and a linear output layer 41 | def get_DAE(nu=2000): 42 | iv_dim = 600 43 | inputs = Input(shape=(iv_dim,)) 44 | x = Dense(nu)(inputs) 45 | x = Activation('tanh')(x) 46 | x = Dense(iv_dim)(x) 47 | out = Activation('linear')(x) 48 | model = Model(inputs=inputs, outputs=out) 49 | 50 | return model 51 | 52 | # Making dictionary to find blacklist pair between train and test dataset 53 | bl_match = np.loadtxt('data/bl_matching.csv',dtype='str') 54 | dev2train={} 55 | dev2id={} 56 | train2dev={} 57 | train2id={} 58 | test2train={} 59 | train2test={} 60 | for iter, line in enumerate(bl_match): 61 | line_s = line.split(',') 62 | dev2train[line_s[1].split('_')[-1]]= line_s[3].split('_')[-1] 63 | dev2id[line_s[1].split('_')[-1]]= line_s[0].split('_')[-1] 64 | train2dev[line_s[3].split('_')[-1]]= line_s[1].split('_')[-1] 65 | train2id[line_s[3].split('_')[-1]]= line_s[0].split('_')[-1] 66 | test2train[line_s[2].split('_')[-1]]= line_s[3].split('_')[-1] 67 | train2test[line_s[3].split('_')[-1]]= line_s[2].split('_')[-1] 68 | 69 | 70 | # load test set information 71 | filename = 'data/tst_evaluation_keys.csv' 72 | tst_info = np.loadtxt(filename,dtype='str',delimiter=',',skiprows=1,usecols=range(0,3)) 73 | tst_trials = [] 74 | tst_trials_label = [] 75 | tst_ground_truth =[] 76 | for iter in range(len(tst_info)): 77 | tst_trials_label.extend([tst_info[iter,0]]) 78 | if tst_info[iter,1]=='background': 79 | tst_trials = np.append(tst_trials,0) 80 | 81 | else: 82 | tst_trials = np.append(tst_trials,1) 83 | 84 | 85 | # Set random seed to make results reproducible 86 | seed = 134 87 | np.random.seed(seed) 88 | 89 | # Loading i-vector 90 | trn_bl_id, trn_bl_utt, trn_bl_ivector = load_ivector('data/trn_blacklist.csv') 91 | trn_bg_id, trn_bg_utt, trn_bg_ivector = load_ivector('data/trn_background.csv') 92 | dev_bl_id, dev_bl_utt, dev_bl_ivector = load_ivector('data/dev_blacklist.csv') 93 | dev_bg_id, dev_bg_utt, dev_bg_ivector = load_ivector('data/dev_background.csv') 94 | tst_id, test_utt, tst_ivector = load_ivector('data/tst_evaluation.csv') 95 | 96 | # length normalization 97 | trn_bl_ivector = length_norm(trn_bl_ivector) 98 | trn_bg_ivector = length_norm(trn_bg_ivector) 99 | dev_bl_ivector = length_norm(dev_bl_ivector) 100 | dev_bg_ivector = length_norm(dev_bg_ivector) 101 | tst_ivector = length_norm(tst_ivector) 102 | 103 | # Inputs to DAE are ivectors and targets are the speaker-level mean ivectors 104 | train_spk_ids = pd.DataFrame({'spk_ids': trn_bg_id}) 105 | train_ivs = pd.DataFrame(trn_bg_ivector) 106 | 107 | X_train = train_ivs.values 108 | Y_train = (train_ivs.groupby(train_spk_ids['spk_ids']).transform('mean')).values 109 | 110 | # DAE training 111 | model = get_DAE() 112 | 113 | model.compile(loss='cosine_proximity', 114 | optimizer = optimizers.Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=1e-08, decay=1e-06), 115 | metrics=[metrics.mean_squared_error]) 116 | 117 | num_examples = X_train.shape[0] 118 | num_epochs = 5 119 | batch_size = 512 120 | num_batch_per_epoch = num_examples / batch_size 121 | 122 | model.fit(x=X_train,y=Y_train,batch_size=batch_size,epochs=num_epochs) 123 | 124 | # Compute DAE-transformed embeddings from ivectors 125 | trn_bl_embeddings = model.predict(trn_bl_ivector,batch_size=batch_size) 126 | trn_bg_embeddings = model.predict(trn_bg_ivector,batch_size=batch_size) 127 | dev_bl_embeddings = model.predict(dev_bl_ivector,batch_size=batch_size) 128 | dev_bg_embeddings = model.predict(dev_bg_ivector,batch_size=batch_size) 129 | tst_embeddings = model.predict(tst_ivector,batch_size=batch_size) 130 | 131 | # Calculating speaker mean vector 132 | spk_mean, spk_mean_label = make_spkvec(trn_bl_embeddings,trn_bl_id) 133 | 134 | # length normalization 135 | trn_bl_embeddings = length_norm(trn_bl_embeddings) 136 | trn_bg_embeddings = length_norm(trn_bg_embeddings) 137 | dev_bl_embeddings = length_norm(dev_bl_embeddings) 138 | dev_bg_embeddings = length_norm(dev_bg_embeddings) 139 | tst_embeddings = length_norm(tst_embeddings) 140 | 141 | print('Dev set score using train set :') 142 | 143 | # making trials of Dev set 144 | dev_embeddings = np.append(dev_bl_embeddings, dev_bg_embeddings,axis=0) 145 | dev_trials = np.append( np.ones([len(dev_bl_id), 1]), np.zeros([len(dev_bg_id), 1])) 146 | 147 | # Cosine distance scoring 148 | scores = spk_mean.dot(dev_embeddings.transpose()) 149 | dev_scores = np.max(scores,axis=0) 150 | 151 | # Top-S detector EER 152 | dev_EER = calculate_EER(dev_trials, dev_scores) 153 | 154 | #divide trial label into target and non-target, plus confusion error(blacklist, fail at blacklist detector) 155 | dev_identified_label = spk_mean_label[np.argmax(scores,axis=0)] 156 | dev_trials_label = np.append( dev_bl_id,dev_bg_id) 157 | 158 | # Top-1 detector EER 159 | dev_trials_confusion = get_trials_label_with_confusion(dev_identified_label, dev_trials_label, dev2train, dev_trials ) 160 | dev_EER_confusion = calculate_EER_with_confusion(dev_scores,dev_trials_confusion) 161 | 162 | print('Test set score using train set:') 163 | 164 | #Cosine distance scoring on Test set 165 | scores = spk_mean.dot(tst_embeddings.transpose()) 166 | tst_scores = np.max(scores,axis=0) 167 | 168 | # top-S detector EER 169 | tst_EER = calculate_EER(tst_trials, tst_scores) 170 | 171 | #divide trial label into target and non-target, plus confusion error(blacklist, fail at blacklist detector) 172 | tst_identified_label = spk_mean_label[np.argmax(scores,axis=0)] 173 | 174 | # Top-1 detector EER 175 | tst_trials_confusion = get_trials_label_with_confusion(tst_identified_label, tst_trials_label, test2train, tst_trials ) 176 | tst_EER_confusion = calculate_EER_with_confusion(tst_scores,tst_trials_confusion) 177 | 178 | 179 | print('Test set score using train + dev set:') 180 | 181 | # get dev set id consistent with Train set 182 | dev_bl_id_along_trnset = [] 183 | for iter in range(len(dev_bl_id)): 184 | dev_bl_id_along_trnset.extend([dev2train[dev_bl_id[iter]]]) 185 | 186 | # Calculating speaker mean vector 187 | spk_mean, spk_mean_label = make_spkvec(np.append(trn_bl_embeddings,dev_bl_embeddings,0),np.append(trn_bl_id,dev_bl_id_along_trnset)) 188 | 189 | #Cosine distance scoring on Test set 190 | scores = spk_mean.dot(tst_embeddings.transpose()) 191 | tst_scores = np.max(scores,axis=0) 192 | 193 | # top-S detector EER 194 | tst_EER = calculate_EER(tst_trials, tst_scores) 195 | 196 | #divide trial label into target and non-target, plus confusion error(blacklist, fail at blacklist detector) 197 | tst_identified_label = spk_mean_label[np.argmax(scores,axis=0)] 198 | 199 | # Top-1 detector EER 200 | tst_trials_confusion = get_trials_label_with_confusion(tst_identified_label, tst_trials_label, test2train,tst_trials ) 201 | tst_EER_confusion = calculate_EER_with_confusion(tst_scores,tst_trials_confusion) 202 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "{}" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright {yyyy} {name of copyright owner} 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. --------------------------------------------------------------------------------