ResNeSt

Last updated on Feb 23, 2021

Cascade Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch)

Memory (M) 10500.0
Backbone Layers 101
File Size 397.39 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, Dense Connections, FPN, RoIAlign
lr sched 1x
Memory (M) 10500.0
Backbone Layers 101
SHOW MORE
SHOW LESS
Cascade Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch)

lr sched 1x
Backbone Layers 50
File Size 317.76 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, Dense Connections, FPN, RoIAlign
lr sched 1x
Backbone Layers 50
SHOW MORE
SHOW LESS
Cascade R-CNN ResNeSt (S-101-FPN, 1x, pytorch)

Memory (M) 8400.0
Backbone Layers 101
File Size 367.09 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture RPN, ResNeSt, FPN, Cascade R-CNN, RoIAlign
lr sched 1x
Memory (M) 8400.0
Backbone Layers 101
SHOW MORE
SHOW LESS
Cascade R-CNN ResNeSt (S-50-FPN, 1x, pytorch)

lr sched 1x
Backbone Layers 50
File Size 287.46 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture RPN, ResNeSt, FPN, Cascade R-CNN, RoIAlign
lr sched 1x
Backbone Layers 50
SHOW MORE
SHOW LESS
Faster R-CNN ResNeSt (S-101-FPN, 1x, pytorch)

Memory (M) 7100.0
Backbone Layers 101
File Size 251.61 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, FPN, RoIPool
lr sched 1x
Memory (M) 7100.0
Backbone Layers 101
SHOW MORE
SHOW LESS
Faster R-CNN ResNeSt (S-50-FPN, 1x, pytorch)

Memory (M) 4800.0
Backbone Layers 50
File Size 171.97 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, FPN, RoIPool
lr sched 1x
Memory (M) 4800.0
Backbone Layers 50
SHOW MORE
SHOW LESS
Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch)

Memory (M) 7800.0
Backbone Layers 101
File Size 261.71 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, Dense Connections, FPN, RoIAlign
lr sched 1x
Memory (M) 7800.0
Backbone Layers 101
SHOW MORE
SHOW LESS
Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch)

Memory (M) 5500.0
Backbone Layers 50
File Size 182.07 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time

Architecture Softmax, RPN, ResNeSt, Convolution, Dense Connections, FPN, RoIAlign
lr sched 1x
Memory (M) 5500.0
Backbone Layers 50
SHOW MORE
SHOW LESS
README.md

ResNeSt: Split-Attention Networks

Introduction

[BACKBONE]

@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}
}

Results and Models

Faster R-CNN

Backbone Style Lr schd Mem (GB) Inf time (fps) box AP Config Download
S-50-FPN pytorch 1x 4.8 - 42.0 config model | log
S-101-FPN pytorch 1x 7.1 - 44.5 config model | log

Mask R-CNN

Backbone Style Lr schd Mem (GB) Inf time (fps) box AP mask AP Config Download
S-50-FPN pytorch 1x 5.5 - 42.6 38.1 config model | log
S-101-FPN pytorch 1x 7.8 - 45.2 40.2 config model | log

Cascade R-CNN

Backbone Style Lr schd Mem (GB) Inf time (fps) box AP Config Download
S-50-FPN pytorch 1x - - 44.5 config model | log
S-101-FPN pytorch 1x 8.4 - 46.8 config model | log

Cascade Mask R-CNN

Backbone Style Lr schd Mem (GB) Inf time (fps) box AP mask AP Config Download
S-50-FPN pytorch 1x - - 45.4 39.5 config model | log
S-101-FPN pytorch 1x 10.5 - 47.7 41.4 config model | log

Results

Object Detection on COCO minival
MODEL BOX AP
Cascade Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 47.7
Cascade R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 46.8
Cascade Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 45.4
Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 45.2
Faster R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 44.5
Cascade R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 44.5
Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 42.6
Faster R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 42.0
Instance Segmentation on COCO minival
MODEL MASK AP
Cascade Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 41.4
Mask R-CNN ResNeSt (S-101-FPN, 1x, pytorch) 40.2
Cascade Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 39.5
Mask R-CNN ResNeSt (S-50-FPN, 1x, pytorch) 38.1