Skip to content

Latest commit

 

History

History

bisenetv1

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

BiSeNetV1

BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation

Introduction

Official Repo

Code Snippet

Abstract

Semantic segmentation requires both rich spatial information and sizeable receptive field. However, modern approaches usually compromise spatial resolution to achieve real-time inference speed, which leads to poor performance. In this paper, we address this dilemma with a novel Bilateral Segmentation Network (BiSeNet). We first design a Spatial Path with a small stride to preserve the spatial information and generate high-resolution features. Meanwhile, a Context Path with a fast downsampling strategy is employed to obtain sufficient receptive field. On top of the two paths, we introduce a new Feature Fusion Module to combine features efficiently. The proposed architecture makes a right balance between the speed and segmentation performance on Cityscapes, CamVid, and COCO-Stuff datasets. Specifically, for a 2048x1024 input, we achieve 68.4% Mean IOU on the Cityscapes test dataset with speed of 105 FPS on one NVIDIA Titan XP card, which is significantly faster than the existing methods with comparable performance.

Results and models

Cityscapes

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) Device mIoU mIoU(ms+flip) config download
BiSeNetV1 R-18-D32 (No Pretrain) 1024x1024 160000 5.69 31.77 V100 74.44 77.05 config model | log
BiSeNetV1 R-18-D32 1024x1024 160000 5.69 31.77 V100 74.37 76.91 config model | log
BiSeNetV1 R-18-D32 (4x8) 1024x1024 160000 11.17 31.77 V100 75.16 77.24 config model | log
BiSeNetV1 R-50-D32 (No Pretrain) 1024x1024 160000 15.39 7.71 V100 76.92 78.87 config model | log
BiSeNetV1 R-50-D32 1024x1024 160000 15.39 7.71 V100 77.68 79.57 config model | log

COCO-Stuff 164k

Method Backbone Crop Size Lr schd Mem (GB) Inf time (fps) Device mIoU mIoU(ms+flip) config download
BiSeNetV1 R-18-D32 (No Pretrain) 512x512 160000 - - V100 25.45 26.15 config model | log
BiSeNetV1 R-18-D32 512x512 160000 6.33 74.24 V100 28.55 29.26 config model | log
BiSeNetV1 R-50-D32 (No Pretrain) 512x512 160000 - - V100 29.82 30.33 config model | log
BiSeNetV1 R-50-D32 512x512 160000 9.28 32.60 V100 34.88 35.37 config model | log
BiSeNetV1 R-101-D32 (No Pretrain) 512x512 160000 - - V100 31.14 31.76 config model | log
BiSeNetV1 R-101-D32 512x512 160000 10.36 25.25 V100 37.38 37.99 config model | log

Note:

  • 4x8: Using 4 GPUs with 8 samples per GPU in training.
  • For BiSeNetV1 on Cityscapes dataset, default setting is 4 GPUs with 4 samples per GPU in training.
  • No Pretrain means the model is trained from scratch.

Citation

@inproceedings{yu2018bisenet,
  title={Bisenet: Bilateral segmentation network for real-time semantic segmentation},
  author={Yu, Changqian and Wang, Jingbo and Peng, Chao and Gao, Changxin and Yu, Gang and Sang, Nong},
  booktitle={Proceedings of the European conference on computer vision (ECCV)},
  pages={325--341},
  year={2018}
}