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) |