yolo3
- extract frames from sample video (frame_extract.py)
- label bounding boxes using boobs.json
- augment image set with imgaug -- at least 2000 images (image_augment.py)
- train yolo3 model (train.py)
- run test video through model (yolo_video.py)
Prerequisites
https://github.com/qqwweee/keras-yolo3
https://github.com/aleju/imgaug
https://github.com/drainingsun/boobs