GroupNorm

open-mmlab / mmdetection

Last updated on Feb 23, 2021
Mask R-CNN GroupNorm (R-101-FPN (d), 2x)
Memory (M) 9900.0
inference time (s/im) 0.11111
File Size 246.89 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 2x
Memory (M) 9900.0
Backbone Layers 101
inference time (s/im) 0.11111
SHOW MORE
SHOW LESS
Mask R-CNN GroupNorm (R-101-FPN (d), 3x)
Memory (M) 9900.0
Backbone Layers 101
File Size 246.89 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 3x
Memory (M) 9900.0
Backbone Layers 101
SHOW MORE
SHOW LESS
Mask R-CNN GroupNorm (R-50-FPN (c), 2x)
Memory (M) 7100.0
inference time (s/im) 0.09174
File Size 174.42 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 2x
Memory (M) 7100.0
Backbone Layers 50
inference time (s/im) 0.09174
SHOW MORE
SHOW LESS
Mask R-CNN GroupNorm (R-50-FPN (c), 3x)
Memory (M) 7100.0
Backbone Layers 50
File Size 174.42 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 3x
Memory (M) 7100.0
Backbone Layers 50
SHOW MORE
SHOW LESS
Mask R-CNN GroupNorm (R-50-FPN (d), 2x)
Memory (M) 7100.0
inference time (s/im) 0.09091
File Size 174.42 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 2x
Memory (M) 7100.0
Backbone Layers 50
inference time (s/im) 0.09091
SHOW MORE
SHOW LESS
Mask R-CNN GroupNorm (R-50-FPN (d), 3x)
Memory (M) 7100.0
Backbone Layers 50
File Size 174.42 MB
Training Data MS COCO
Training Resources 8x NVIDIA V100 GPUs
Training Time
Architecture Softmax, RPN, Convolution, Dense Connections, Group Normalization, FPN, ResNet, RoIAlign
lr sched 3x
Memory (M) 7100.0
Backbone Layers 50
SHOW MORE
SHOW LESS
README.md

Group Normalization

Introduction

[ALGORITHM]

@inproceedings{wu2018group,
  title={Group Normalization},
  author={Wu, Yuxin and He, Kaiming},
  booktitle={Proceedings of the European Conference on Computer Vision (ECCV)},
  year={2018}
}

Results and Models

Backbone model Lr schd Mem (GB) Inf time (fps) box AP mask AP Config Download
R-50-FPN (d) Mask R-CNN 2x 7.1 11.0 40.2 36.4 config model | log
R-50-FPN (d) Mask R-CNN 3x 7.1 - 40.5 36.7 config model | log
R-101-FPN (d) Mask R-CNN 2x 9.9 9.0 41.9 37.6 config model | log
R-101-FPN (d) Mask R-CNN 3x 9.9 42.1 38.0 config model | log
R-50-FPN (c) Mask R-CNN 2x 7.1 10.9 40.0 36.1 config model | log
R-50-FPN (c) Mask R-CNN 3x 7.1 - 40.1 36.2 config model | log

Notes:

  • (d) means pretrained model converted from Detectron, and (c) means the contributed model pretrained by @thangvubk.
  • The 3x schedule is epoch [28, 34, 36].
  • Memory, Train/Inf time is outdated.

Results

Object Detection on COCO minival

Object Detection on COCO minival
MODEL BOX AP
Mask R-CNN GroupNorm (R-101-FPN (d), 3x) 42.1
Mask R-CNN GroupNorm (R-101-FPN (d), 2x) 41.9
Mask R-CNN GroupNorm (R-50-FPN (d), 3x) 40.5
Mask R-CNN GroupNorm (R-50-FPN (d), 2x) 40.2
Mask R-CNN GroupNorm (R-50-FPN (c), 3x) 40.1
Mask R-CNN GroupNorm (R-50-FPN (c), 2x) 40.0
Instance Segmentation on COCO minival
MODEL MASK AP
Mask R-CNN GroupNorm (R-101-FPN (d), 3x) 38.0
Mask R-CNN GroupNorm (R-101-FPN (d), 2x) 37.6
Mask R-CNN GroupNorm (R-50-FPN (d), 3x) 36.7
Mask R-CNN GroupNorm (R-50-FPN (d), 2x) 36.4
Mask R-CNN GroupNorm (R-50-FPN (c), 3x) 36.2
Mask R-CNN GroupNorm (R-50-FPN (c), 2x) 36.1
Contact us on: hello@paperswithcode.com . Papers With Code is a free resource with all data licensed under CC-BY-SA.
Terms Data policy Cookies policy