In this project, we are interested in building a pedestrian detector. We train a Faster r-cnn network on the Penn-Fudan Database for Pedestrian Detection. The Penn-Fudan Database images are taken from scenes around campus and urban street. The objects we are interested in these images are pedestrians. Each image will have at least one pedestrian in it.

• Python (3.6)
• PyTorch deep learning framework(1.2.0)
• Torchvision (0.4.0)
• The training was done on an GPU Nvidia GTX1050

First, download the Penn-Fudan database from here, unzip it and place it on folder data.

To train the model, navigate (cd) to .\code\and run

python train.py

The weights will be saved to .\model\$

The model can be tested by running

python test.py IMAGE_PATH