ResNeSt: Split-Attention Networks

This is an implementation of "ResNeSt : Split-Attention Networks" on Keras and Tensorflow.

The implementation is based on paper[1] and official implementation.

Model

• Model
  • ResNeSt-50
  • ResNeSt-101
  • ResNeSt-200
  • ResNeSt-269
• Pre-trained weight
  • Imagenet(The pre-trained weights are converted from official weight.)

Requirements

• Python 3
• tensorflow 2
• torch 1.1▲ (Use the pre-trained weights)

Reference

• ResNeSt : Split-Attention Networks, Hang Zhang, Chongruo Wu, Zhongyue Zhang, Yi Zhu, Zhi Zhang, Haibin Lin, Yue Sun, Tong He, Jonas Mueller, R. Manmatha, Mu Li, Alexander Smola, https://arxiv.org/abs/2004.08955

Contributor

• Hyungjin Kim(flslzk@gmail.com)