# Improving Semantic Segmentation via Decoupled Body and Edge Supervision

@article{Li2020ImprovingSS,
title={Improving Semantic Segmentation via Decoupled Body and Edge Supervision},
author={Xiangtai Li and Xia Li and Li Zhang and Guangliang Cheng and Jianping Shi and Zhouchen Lin and Shaohua Tan and Yunhai Tong},
journal={ArXiv},
year={2020},
volume={abs/2007.10035}
}
Existing semantic segmentation approaches either aim to improve the object's inner consistency by modeling the global context, or refine objects detail along their boundaries by multi-scale feature fusion. In this paper, a new paradigm for semantic segmentation is proposed. Our insight is that appealing performance of semantic segmentation requires \textit{explicitly} modeling the object \textit{body} and \textit{edge}, which correspond to the high and low frequency of the image. To do so, we… Expand

#### Figures and Tables from this paper

Dynamic Dual Sampling Module for Fine-Grained Semantic Segmentation
• Chen Shi, Yanran Wu
• Computer Science
• ICIP 2021
• 2021
A Dynamic Dual Sampling Module (DDSM) is proposed to conduct dynamic affinity modeling and propagate semantic context to local details, which yields a more discriminative representation. Expand
Enhanced Boundary Learning for Glass-like Object Segmentation
A novel refined differential module for generating finer boundary cues and an edge-aware pointbased graph convolution network module is proposed to model the global shape representation along the boundary to solve the glass-like object segmentation problem via enhanced boundary learning. Expand
MDNet: A Multi-image Input Real-Time Semantic Segmentation Model with Decoupled Supervision
• Yunze Wu
• 2021 the 5th International Conference on Information System and Data Mining
• 2021
Current state-of-the-art Real-Time Semantic Segmentation Model is still not fast enough. They spend too much time on processing images in a deep CNN to grab the spatial and context information.Expand
Active Boundary Loss for Semantic Segmentation
Experimental results show that training with the active boundary loss can effectively improve the boundary F-score and mean Intersection-over-Union on challenging image and video object segmentation datasets. Expand
A Unified Efficient Pyramid Transformer for Semantic Segmentation
• Fangrui Zhu, Li Zhang, Yanwei Fu
• Computer Science
• ArXiv
• 2021
This work advocate a unified framework (UN-EPT) to segment objects by considering both context information and boundary artifacts by adapting a sparse sampling strategy to incorporate the transformer-based attention mechanism for efficient context modeling. Expand
Semantic-Guided Representation Enhancement for Self-supervised Monocular Trained Depth Estimation
A semantic-guided depth representation enhancement method, which promotes both local and global depth feature representations by leveraging rich contextual information and outperforms the state-of-the-art methods. Expand
BoundarySqueeze: Image Segmentation as Boundary Squeezing
The proposed Boundary Squeeze module is a novel and efficient module that squeezes the object boundary from both inner and outer directions which leads to precise mask representation and outperforms previous state-of-theart PointRend in both accuracy and speed under the same setting. Expand
A Survey on Deep Learning Methods for Semantic Image Segmentation in Real-Time
• G. Takos
• Computer Science, Mathematics
• ArXiv
• 2020
This work provides a comprehensive analysis of state-of-the-art deep learning architectures in image segmentation and an extensive list of techniques to achieve fast inference and computational efficiency. Expand
ME-Net: Multi-Step Enhancement Network for Semantic Segmentation
• Xingming Zhang, Shiliang Huang, Jun Yin
• 2021 6th International Conference on Multimedia Systems and Signal Processing
• 2021
∗Semantic segmentation faces two problems of one is that the detail information of great significance for pixel classification is insufficient, and of another one is the contextual information isExpand
Deep Interactive Thin Object Selection
• Computer Science
• 2021 IEEE Winter Conference on Applications of Computer Vision (WACV)
• 2021
The effectiveness of the proposed solution on segmenting thin objects is demonstrated, surpassing the baseline by ~ 30% IoUthin despite using only four clicks, and a novel integrative thin object segmentation network consisting of three streams. Expand

#### References

SHOWING 1-10 OF 73 REFERENCES
Semantic Segmentation with Boundary Neural Fields
• Computer Science, Mathematics
• 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
• 2016
A Boundary Neural Field (BNF) is introduced, which is a global energy model integrating FCN predictions with boundary cues that is used to enhance semantic segment coherence and to improve object localization. Expand
Improving Semantic Segmentation via Video Propagation and Label Relaxation
• Yi Zhu, +4 authors Bryan Catanzaro
• Computer Science
• 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
• 2019
This paper presents a video prediction-based methodology to scale up training sets by synthesizing new training samples in order to improve the accuracy of semantic segmentation networks, and introduces a novel boundary label relaxation technique that makes training robust to annotation noise and propagation artifacts along object boundaries. Expand
Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform
• Computer Science
• 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
• 2016
This work proposes replacing the fully-connected CRF with domain transform (DT), a modern edge-preserving filtering method in which the amount of smoothing is controlled by a reference edge map, and shows that it yields comparable semantic segmentation results, accurately capturing object boundaries. Expand
Adaptive Affinity Fields for Semantic Segmentation
• Computer Science
• ECCV
• 2018
This work proposes the concept of Adaptive Affinity Fields (AAF), a simpler alternative that learns to verify the spatial structure of segmentation during training only, and uses adversarial learning to select the optimal affinity field size for each semantic category. Expand
Boundary-Aware Feature Propagation for Scene Segmentation
• Computer Science
• 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
• 2019
A boundary-aware feature propagation (BFP) module to harvest and propagate the local features within their regions isolated by the learned boundaries in the UAG-structured image and achieves new state-of-the-art segmentation performance on three challenging semantic segmentation datasets, i.e., PASCAL-Context, CamVid, and Cityscapes. Expand
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation
• Computer Science
• 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
• 2016
This work shows how to improve semantic segmentation through the use of contextual information, specifically, ' patch-patch' context between image regions, and 'patch-background' context, and formulate Conditional Random Fields with CNN-based pairwise potential functions to capture semantic correlations between neighboring patches. Expand
Convolutional Random Walk Networks for Semantic Image Segmentation
• Computer Science
• 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
• 2017
This work introduces a simple, yet effective Convolutional Random Walk Network (RWN) that addresses the issues of poor boundary localization and spatially fragmented predictions with very little increase in model complexity. Expand
Dual Attention Network for Scene Segmentation
• Computer Science
• 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
• 2019
New state-of-the-art segmentation performance on three challenging scene segmentation datasets, i.e., Cityscapes, PASCAL Context and COCO Stuff dataset is achieved without using coarse data. Expand
Dual Graph Convolutional Network for Semantic Segmentation
• Computer Science
• BMVC
• 2019
The Dual Graph Convolutional Network (DGCNet) models the global context of the input feature by modelling two orthogonal graphs in a single framework, which achieves state-of-the-art results on both Cityscapes and Pascal Context datasets. Expand
Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
• Computer Science
• ICLR
• 2015
This work brings together methods from DCNNs and probabilistic graphical models for addressing the task of pixel-level classification by combining the responses at the final DCNN layer with a fully connected Conditional Random Field (CRF). Expand