Adversarial Multi Instance Learning for Human Pose Estimation

AMIL Architecture

Python Implementation of ["AMIL: Adversarial Multi Instance Learning for Human Pose Estimation"]

Results

Prepare Data

• 1. The MPII and PoseTrack dataset are available in data directory.
• 2. need to have the high resolution images for training.
  • In this experiment, we used images from [MPII], so the hyper-paremeters in config.py (like number of epochs) are seleted basic on that dataset, if you change a larger dataset you can reduce the number of epochs.
  • If you dont want to use these dataset, you can also use COCO, just simply download it using train_hr_imgs = tl.files.load_dataset_coco2017 in main.py.

Citation

If you find this project useful, we would be grateful if you cite the paper：

@article{TOMM 2019,
author = {Pourya Shamsolmoali, Masoumeh Zareapoor, Huiyu Zhou and Jie Yang},
journal = {ACM Transactions on Multimedia Computing Communications and Applications},
title = {{AMIL: Adversarial Multi-Instance Learning for Human Pose Estimation}},
year = {2019}
}