You can find our project report at report.pdf
Under pose/models/hourglass.py you can find our improvements.
To run an experiment:
python ./example/main.py --dataset mpii --arch hg --stack 1 --block 1 --features 256 --checkpoint checkpoint/s1/
notice that the MPII dataset must be located in /data/mpii/
checkpoint/ - saves the model after each epoch when increasing accuracy data/ - should contain the datasets pose/ - the model's code
PyTorch (>= 0.4.1): Please follow the installation instruction of PyTorch.
Download annotation file:
(MPII) Download mpii_annotations.json and save it to data/mpii