Skip to content

sunke123/ResNeSt

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PyPI PyPI Pre-release PyPI Nightly Downloads License Unit Test arXiv

PWC PWC PWC PWC PWC PWC

ResNeSt

Split-Attention Network, A New ResNet Variant. It significantly boosts the performance of downstream models such as Mask R-CNN, Cascade R-CNN and DeepLabV3.

Table of Contents

  1. Pretrained Models
  2. Transfer Learning Models
  3. Verify Backbone Models
  4. How to Train
  5. Reference

Pypi / GitHub Install

  1. Install this package repo, note that you only need to choose one of the options
# using github url
pip install git+https://github.com/zhanghang1989/ResNeSt

# using pypi
pip install resnest --pre

Pretrained Models

crop size PyTorch Gluon
ResNeSt-50 224 81.03 81.04
ResNeSt-101 256 82.83 82.81
ResNeSt-200 320 83.84 83.88
ResNeSt-269 416 84.54 84.53
  • 3rd party implementations are available: Tensorflow, Caffe.

  • Extra ablation study models are available in link

PyTorch Models

  • Load using Torch Hub
import torch
# get list of models
torch.hub.list('zhanghang1989/ResNeSt', force_reload=True)

# load pretrained models, using ResNeSt-50 as an example
net = torch.hub.load('zhanghang1989/ResNeSt', 'resnest50', pretrained=True)
  • Load using python package
# using ResNeSt-50 as an example
from resnest.torch import resnest50
net = resnest50(pretrained=True)

Gluon Models

  • Load pretrained model:
# using ResNeSt-50 as an example
from resnest.gluon import resnest50
net = resnest50(pretrained=True)

Transfer Learning Models

Detectron Models

Training code and pretrained models are released at our Detectron2 Fork.

Object Detection on MS-COCO validation set

Method Backbone mAP%
Faster R-CNN ResNet-50 39.25
ResNet-101 41.37
ResNeSt-50 (ours) 42.33
ResNeSt-101 (ours) 44.72
Cascade R-CNN ResNet-50 42.52
ResNet-101 44.03
ResNeSt-50 (ours) 45.41
ResNeSt-101 (ours) 47.50
ResNeSt-200 (ours) 49.03

Instance Segmentation

Method Backbone bbox mask
Mask R-CNN ResNet-50 39.97 36.05
ResNet-101 41.78 37.51
ResNeSt-50 (ours) 42.81 38.14
ResNeSt-101 (ours) 45.75 40.65
Cascade R-CNN ResNet-50 43.06 37.19
ResNet-101 44.79 38.52
ResNeSt-50 (ours) 46.19 39.55
ResNeSt-101 (ours) 48.30 41.56
ResNeSt-200 (w/ tricks ours) 50.54 44.21
ResNeSt-200-dcn (w/ tricks ours) 50.91 44.50
53.30* 47.10*

All of results are reported on COCO-2017 validation dataset. The values with * demonstrate the mutli-scale testing performance on the test-dev2019.

Panoptic Segmentation

Backbone bbox mask PQ
ResNeSt-200 51.00 43.68 47.90

Semantic Segmentation

Results on ADE20K

Method Backbone pixAcc% mIoU%
Deeplab-V3
ResNet-50 80.39 42.1
ResNet-101 81.11 44.14
ResNeSt-50 (ours) 81.17 45.12
ResNeSt-101 (ours) 82.07 46.91
ResNeSt-200 (ours) 82.45 48.36
ResNeSt-269 (ours) 82.62 47.60

Results on Cityscapes

Method Backbone Split w Mapillary mIoU%
Deeplab-V3+
ResNeSt-200 (ours) Validation no 82.7
ResNeSt-200 (ours) Validation yes 83.8
ResNeSt-200 (ours) Test yes 83.3

Verify Backbone Models:

Note: the inference speed reported in the paper are tested using Gluon implementation with RecordIO data.

Prepare ImageNet dataset:

Here we use raw image data format for simplicity, please follow GluonCV tutorial if you would like to use RecordIO format.

cd scripts/dataset/
# assuming you have downloaded the dataset in the current folder
python prepare_imagenet.py --download-dir ./

Torch Model

# use resnest50 as an example
cd scripts/torch/
python verify.py --model resnest50 --crop-size 224

Gluon Model

# use resnest50 as an example
cd scripts/gluon/
python verify.py --model resnest50 --crop-size 224

How to Train

ImageNet Models

Detectron Models

For object detection and instance segmentation models, please visit our detectron2-ResNeSt fork.

Semantic Segmentation

Reference

ResNeSt: Split-Attention Networks [arXiv]

Hang Zhang, Chongruo Wu, Zhongyue Zhang, Yi Zhu, Zhi Zhang, Haibin Lin, Yue Sun, Tong He, Jonas Muller, R. Manmatha, Mu Li and Alex Smola

@article{zhang2020resnest,
title={ResNeSt: Split-Attention Networks},
author={Zhang, Hang and Wu, Chongruo and Zhang, Zhongyue and Zhu, Yi and Zhang, Zhi and Lin, Haibin and Sun, Yue and He, Tong and Muller, Jonas and Manmatha, R. and Li, Mu and Smola, Alexander},
journal={arXiv preprint arXiv:2004.08955},
year={2020}
}

Major Contributors

About

ResNeSt: Split-Attention Networks

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%