Pytorch implementation of Faster R-CNN for real-time object detection (paper). Our detailed project report is available here.

1. Abhinav Garg ([email protected])
2. Refik Mert Cam ([email protected])
3. Sanyukta Deshpande ([email protected])

Here is an example of create environ from scratch with anaconda

# create conda env
conda create --name frcnn python=3.7
conda activate frcnn
# install pytorch
conda install pytorch torchvision cudatoolkit=10.2 -c pytorch

# install other dependancy
pip install visdom scikit-image tqdm fire ipdb pprint matplotlib torchnet

1. Download the training, validation, test data and VOCdevkit

   wget http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCtrainval_06-Nov-2007.tar
   wget http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCtest_06-Nov-2007.tar
   wget http://host.robots.ox.ac.uk/pascal/VOC/voc2007/VOCdevkit_08-Jun-2007.tar

2. Extract all of these tars into one directory named VOCdevkit

   tar xvf VOCtrainval_06-Nov-2007.tar
   tar xvf VOCtest_06-Nov-2007.tar
   tar xvf VOCdevkit_08-Jun-2007.tar

3. It should have this basic structure

   $VOCdevkit/                           # development kit
   $VOCdevkit/VOCcode/                   # VOC utility code
   $VOCdevkit/VOC2007                    # image sets, annotations, etc.
   # ... and several other directories ...

4. modify voc_data_dir and voc_test_dir cfg item in config/config.py.

Update the parameters in config/config.py as per the experiment. Update save_path to the path where model files are to be stored. Note: Our implementation currently only supports vgg16 and resnet101 for pretrained_model cfg item.

python approx_train.py 

To run inference on select test images, update train=False and save_path to the path where trained model is located in config/config.py.

python test.py