├── object-detection-1.png ├── object-detection-2.png ├── object-detection-3.png ├── object-detection-4.png ├── object-detection-5.png ├── .gitignore └── README.md /object-detection-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ildoonet/deep-object-detection-models/HEAD/object-detection-1.png -------------------------------------------------------------------------------- /object-detection-2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ildoonet/deep-object-detection-models/HEAD/object-detection-2.png -------------------------------------------------------------------------------- /object-detection-3.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ildoonet/deep-object-detection-models/HEAD/object-detection-3.png -------------------------------------------------------------------------------- /object-detection-4.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ildoonet/deep-object-detection-models/HEAD/object-detection-4.png -------------------------------------------------------------------------------- /object-detection-5.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ildoonet/deep-object-detection-models/HEAD/object-detection-5.png -------------------------------------------------------------------------------- /.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 | env/ 12 | build/ 13 | develop-eggs/ 14 | dist/ 15 | downloads/ 16 | eggs/ 17 | .eggs/ 18 | lib/ 19 | lib64/ 20 | parts/ 21 | sdist/ 22 | var/ 23 | *.egg-info/ 24 | .installed.cfg 25 | *.egg 26 | 27 | # PyInstaller 28 | # Usually these files are written by a python script from a template 29 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 30 | *.manifest 31 | *.spec 32 | 33 | # Installer logs 34 | pip-log.txt 35 | pip-delete-this-directory.txt 36 | 37 | # Unit test / coverage reports 38 | htmlcov/ 39 | .tox/ 40 | .coverage 41 | .coverage.* 42 | .cache 43 | nosetests.xml 44 | coverage.xml 45 | *,cover 46 | .hypothesis/ 47 | 48 | # Translations 49 | *.mo 50 | *.pot 51 | 52 | # Django stuff: 53 | *.log 54 | local_settings.py 55 | 56 | # Flask stuff: 57 | instance/ 58 | .webassets-cache 59 | 60 | # Scrapy stuff: 61 | .scrapy 62 | 63 | # Sphinx documentation 64 | docs/_build/ 65 | 66 | # PyBuilder 67 | target/ 68 | 69 | # IPython Notebook 70 | .ipynb_checkpoints 71 | 72 | # pyenv 73 | .python-version 74 | 75 | # celery beat schedule file 76 | celerybeat-schedule 77 | 78 | # dotenv 79 | .env 80 | 81 | # virtualenv 82 | venv/ 83 | ENV/ 84 | 85 | # Spyder project settings 86 | .spyderproject 87 | 88 | # Rope project settings 89 | .ropeproject 90 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # deep-object-detection-models 2 | Deep Learning으로 학습된 Object Detection Model 에 대해 정리한 Archive 임. 3 | 4 | ## 1편 : [SlideShare](https://www.slideshare.net/IldooKim/deep-object-detectors-1-20166) 5 | 6 | [SlideShare 에서 정리된 슬라이드보기](https://www.slideshare.net/IldooKim/deep-object-detectors-1-20166) 7 | 8 | Deep Learning 기반의 모델링으로 Object Detection 문제를 효과적으로 풀기위해 시도되었던 다양한 내용들을 정리한 자료입니다. 9 | 10 | ![Summary SlideShare #1](/object-detection-1.png?raw=true "Summary") 11 | ![Summary : Before R-CNN](/object-detection-2.png?raw=true "Before R-CNN") 12 | ![Summary : R-CNN](/object-detection-3.png?raw=true "R-CNN") 13 | ![Summary : Single Shot Detector](/object-detection-4.png?raw=true "Single Shot detector 1") 14 | ![Summary : Single Shot Detector](/object-detection-5.png?raw=true "Single Shot detector 2") 15 | 16 | 17 | ### R-CNN 이전의 모델 18 | 19 | * DPM : [Object Detection with Discriminatively TrainedPart Based Models](http://cs.brown.edu/~pff/papers/lsvm-pami.pdf) 20 | * [Selective Search for Object Recognition](http://cs.brown.edu/~pff/papers/lsvm-pami.pdf) 21 | * [OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks](https://arxiv.org/abs/1312.6229) 22 | * [Deep Neural Networks for Object Detection](https://pdfs.semanticscholar.org/713f/73ce5c3013d9fb796c21b981dc6629af0bd5.pdf) 23 | 24 | ### R-CNN 류의 모델 : R-CNN의 모듈들을 개선했거나 유사 구조의 Detection Pipeline을 사용 25 | 26 | * R-CNN : [Rich feature hierarchies for accurate object detection and semantic segmentation](https://arxiv.org/abs/1311.2524) 27 |  * [Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition](https://arxiv.org/abs/1406.4729) : 이미지를 28 |  반복해서 Crop해 DNN Feature를 얻는 과정을 개선하기 위해 SPP Layer소개. 29 |  * [Object detection via a multi-region & semantic segmentation-aware CNN model](http://arxiv.org/abs/1505.01749) : Feature Level   30 |  RNN으로 Context Feature Extract하여 Concat 31 | * [Fast R-CNN](https://arxiv.org/abs/1504.08083) : R-CNN의 파이프라인에서 마지막 부분에 해당하는 SVM을 Neural Network로 개선해 성능과 속도를 높임. 32 | * [Faster R-CNN](http://arxiv.org/abs/1506.01497) : R-CNN의 파이프라인에서 첫번쨰 부분에 해당하는 Selective Search(Proposal)을 Neural Network로 개선해 속도와 성능 개선함. 33 | * [R-CNN Minus R](http://arxiv.org/abs/1506.06981) 34 | * [G-CNN: an Iterative Grid Based Object Detector](http://arxiv.org/abs/1512.07729) 35 | * [Faster R-CNN +++](https://arxiv.org/abs/1512.03385) 36 | * [R-FCN: Object Detection via Region-based Fully Convolutional Networks](https://arxiv.org/abs/1605.06409) : Faster R-CNN을 Fully Convolutional 하게 변경. 37 | 38 | ### Single Shot Detector : 1번의 Neural Network Forwarding으로 여러 클래스의 여러 물체를 동시 검출하는 Pipeline을 사용 39 | 40 | * Multibox, Edgebox, MSC-Multibox 등 DNN 기반의 방식으로 Proposal 개선하기도 함. 41 | * [Scalable Object Detection using Deep Neural Networks](https://arxiv.org/abs/1312.2249) 42 | * [Edge Boxes: Locating Object Proposals from Edges](http://research.microsoft.com/pubs/220569/ZitnickDollarECCV14edgeBoxes.pdf) 43 | * [Scalable, High-Quality Object Detection](http://arxiv.org/abs/1412.1441) 44 | * [DeepBox: Learning Objectness With Convolutional Networks](https://github.com/weichengkuo/DeepBox) 45 | * [You Only Look Once: Unified, Real-Time Object Detection](http://arxiv.org/abs/1506.02640) 46 | * Overfeat, Multibox 등의 Formulation을 발전시키고, Network Architecture의 개선으로 성능 개선 47 | * [SSD: Single Shot MultiBox Detector](http://arxiv.org/abs/1512.02325) 48 | * Faster RCNN 등과 성능이 비등하면서도 YOLO만큼 빠름 49 | 50 | ---- 51 | --------------------------------------------------------------------------------