History

Ma Zerun d1473e4a7f [Dependency] Update mmcv dependency version (#509 ) * Update mmcv dependency version * Add code info in some metafiles		2021-11-02 18:08:30 +08:00
..
README.md	[Feature] Add Res2Net backbone and converted weights. (#465 )	2021-10-20 16:34:22 +08:00
metafile.yml	[Dependency] Update mmcv dependency version (#509 )	2021-11-02 18:08:30 +08:00
res2net50-w14-s8_8xb32_in1k.py	[Feature] Add Res2Net backbone and converted weights. (#465 )	2021-10-20 16:34:22 +08:00
res2net50-w26-s8_8xb32_in1k.py	[Feature] Add Res2Net backbone and converted weights. (#465 )	2021-10-20 16:34:22 +08:00
res2net101-w26-s4_8xb32_in1k.py	[Feature] Add Res2Net backbone and converted weights. (#465 )	2021-10-20 16:34:22 +08:00

README.md

Res2Net: A New Multi-scale Backbone Architecture

Introduction

@article{gao2019res2net,
  title={Res2Net: A New Multi-scale Backbone Architecture},
  author={Gao, Shang-Hua and Cheng, Ming-Ming and Zhao, Kai and Zhang, Xin-Yu and Yang, Ming-Hsuan and Torr, Philip},
  journal={IEEE TPAMI},
  year={2021},
  doi={10.1109/TPAMI.2019.2938758},
}

Pretrain model

The pre-trained models are converted from official repo.

ImageNet 1k

Model	resolution	Params(M)	Flops(G)	Top-1 (%)	Top-5 (%)	Download
Res2Net-50-14w-8s*	224x224	25.06	4.22	78.14	93.85	model
Res2Net-50-26w-8s*	224x224	48.40	8.39	79.20	94.36	model
Res2Net-101-26w-4s*	224x224	45.21	8.12	79.19	94.44	model

Models with * are converted from other repos.