This repo is duplication of jwyang/faster-rcnn.pytorch
C/C++ code(nms, roi pool,align) are removed and easier to study.
Python 3.8.5
Ubuntu 20.04.1 LTS (and Window10)
git clone https://github.com/cuchoco/pytorch_faster_rcnn.git
cd pytorch_faster_rcnn && mkdir data
pip install -r requirements.txt
- PASCAL_VOC 07+12: Please follow the instructions in py-faster-rcnn to prepare VOC datasets.
Actually, you can refer to any others. After downloading the data, creat softlinks in the folder data/.
├── data
│ ├── VOCdevkit2007
│ └── pretrained_model
│ └── resnet101_caffe.pth
We used two pretrained models in our experiments, VGG and ResNet101.
You can download these two models from:
Download them and put them into the data/pretrained_model/
- linux
cd data
git clone https://github.com/pdollar/coco.git
cd cocoapi/PythonAPI
make
- window
cd data
git clone https://github.com/philferriere/cocoapi
cd cocoapi/PythonAPI/
python setup.py build_ext --inplace
lib/model/utils/config.py
374 # yaml_cfg = edict(yaml.load(f))
375 yaml_cfg = edict(yaml.safe_load(f))
and replace lib/pycocotools with data/cocoapi/PythonAPI/pycocotools
python trainval_net.py