ResNext (CVPR'2017) ```bibtex @inproceedings{xie2017aggregated, title={Aggregated residual transformations for deep neural networks}, author={Xie, Saining and Girshick, Ross and Doll{\'a}r, Piotr and Tu, Zhuowen and He, Kaiming}, booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition}, pages={1492--1500}, year={2017} } ```
MPII (CVPR'2014) ```bibtex @inproceedings{andriluka14cvpr, author = {Mykhaylo Andriluka and Leonid Pishchulin and Peter Gehler and Schiele, Bernt}, title = {2D Human Pose Estimation: New Benchmark and State of the Art Analysis}, booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, year = {2014}, month = {June} } ```
Results on MPII val set | Arch | Input Size | Mean | Mean@0.1 | ckpt | log | | :---------------------------------------------------------- | :--------: | :---: | :------: | :---------------------------------------------------------: | :---------------------------------------------------------: | | [pose_resnext_152](/configs/body_2d_keypoint/topdown_heatmap/mpii/td-hm_resnext152_8xb32-210e_mpii-256x256.py) | 256x256 | 0.887 | 0.294 | [ckpt](https://download.openmmlab.com/mmpose/top_down/resnext/resnext152_mpii_256x256-df302719_20200927.pth) | [log](https://download.openmmlab.com/mmpose/top_down/resnext/resnext152_mpii_256x256_20200927.log.json) |