Download COCO dataset, place it into ./data, then run python download_and_convert_data.py to build tf-records. It takes a while.
Download pretrained resnet50 model, wget http://download.tensorflow.org/models/resnet_v1_50_2016_08_28.tar.gz, unzip it, place it into ./data/pretrained_models/
Go to ./libs and run make
run python train/train.py for training
There are certainly some bugs, please report them back, and let's solve them together.
TODO:
ROIAlign
COCO Data Provider
Resnet50
Feature Pyramid Network
Anchor and ROI layer
Mask layer
Speedup anchor layer with cython
Combining all modules together.
Testing and debugging (in progress)
Training / evaluation on COCO
Add image summary to show some results
Converting ResneXt
Training >2 images
Call for contributions
Anything helps this repo, including discussion, testing, promotion and of course your awesome code.