README.md

Squeeze-and-excitation networks

Introduction

[BACKBONE]

@inproceedings{hu2018squeeze,
  title={Squeeze-and-excitation networks},
  author={Hu, Jie and Shen, Li and Sun, Gang},
  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
  pages={7132--7141},
  year={2018}
}

Results and models

2d Human Pose Estimation

Results on COCO val2017 with detector having human AP of 56.4 on COCO val2017 dataset

Arch	Input Size	AP	AP50	AP75	AR	AR50	ckpt	log
pose_seresnet_50	256x192	0.728	0.900	0.809	0.784	0.940	ckpt	log
pose_seresnet_50	384x288	0.748	0.905	0.819	0.799	0.941	ckpt	log
pose_seresnet_101	256x192	0.734	0.904	0.815	0.790	0.942	ckpt	log
pose_seresnet_101	384x288	0.753	0.907	0.823	0.805	0.943	ckpt	log
pose_seresnet_152*	256x192	0.730	0.899	0.810	0.786	0.940	ckpt	log
pose_seresnet_152*	384x288	0.753	0.906	0.823	0.806	0.945	ckpt	log

Note that * means without imagenet pre-training.

Results on MPII val set

Arch	Input Size	Mean	[email protected]	ckpt	log
pose_seresnet_50	256x256	0.885	0.334	ckpt	log
pose_seresnet_101	256x256	0.884	0.336	ckpt	log
pose_seresnet_152*	256x256	0.884	0.329	ckpt	log

Note that * means without imagenet pre-training.