mitmul / chainer-faster-rcnn

Object Detection with Faster R-CNN in Chainer
MIT License
288 stars 87 forks source link
chainer

Faster R-CNN

This repo has been deprecated. Here is the complete codes for training Faster-RCNN on your data and using the pre-trained Faster-RCNN model for new data: ChainerCV

This is an experimental implementation of Faster R-CNN in Chainer based on Ross Girshick's work: py-faster-rcnn codes.

Requirement

Using anaconda is strongly recommended.

Installation of dependencies

pip install numpy
pip install cython
pip install chainer
pip install chainercv
# for python3
conda install -c https://conda.binstar.org/menpo opencv3
# for python2
conda install opencv

For Windows users

There's a known problem in cpu_nms.pyx. But a workaround has been posted here (and see also the issue posted to the original py-faster-rcnn).

Setup

1. Build extensions

python setup.py build_ext -i

Inference

1. Download pre-trained model

if [ ! -d data ]; then mkdir data; fi
curl https://dl.dropboxusercontent.com/u/2498135/faster-rcnn/VGG16_faster_rcnn_final.model?dl=1 -o data/VGG16_faster_rcnn_final.model

NOTE: The model definition in faster_rcnn.py has been changed, so if you already have the older pre-trained model file, please download it again to replace the older one with the new one.

2. Use forward.py

curl -O http://vision.cs.utexas.edu/voc/VOC2007_test/JPEGImages/004545.jpg
python forward.py --img_fn 004545.jpg --gpu 0

--gpu 0 turns on GPU. When you turn off GPU, use --gpu -1 or remove --gpu option.

Layers

Summarization of Faster R-CNN layers used during inference

RPN

The region proposal layer (RPN) is consisted of AnchorTargetLayer and ProposalLayer. RPN takes feature maps from trunk network like VGG-16, and performs 3x3 convolution to it. Then, it applies two independent 1x1 convolutions to the output of the first 3x3 convolution. Resulting outputs are rpn_cls_score and rpn_bbox_pred.

Training

1. Make sure chainercv has been installed

ChainerCV is a utility library enables Chainer to treat various datasets easily. It also provides some transformation utility for data augmentation, and includes some standard algorithms for some comptuer vision tasks. Check the repo to know details. Here I use (VOCDetectionDataset)[http://chainercv.readthedocs.io/en/latest/reference/datasets.html#vocdetectiondataset] of ChainerCV. Anyway, before starting training of FasterRCNN, please install ChainerCV via pip.

pip install chainercv

2. Start training

python train_rpn.py

Faster R-CNN Architecture

Note that it is a visualization of the workflow DURING INFERENCE