Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision

2021/06/02 by Xiaokang Chen, Chen, Xiaokang, Yuhui Yuan +5 · 66 citations
Computer Science · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Video Surveillance and Tracking Methods #cs.CV

paper · pdf · doi:10.48550/arxiv.2106.01226

Accepted by CVPR 2021

openalex publication_date 2021/06/02 · arxiv created 2021/06/04 · arxiv updated 2021/06/07 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we study the semi-supervised semantic segmentation problem via exploring both labeled data and extra unlabeled data. We propose a novel consistency regularization approach, called cross pseudo supervision (CPS). Our approach imposes the consistency on two segmentation networks perturbed with different initialization for the same input image. The pseudo one-hot label map, output from one perturbed segmentation network, is used to supervise the other segmentation network with the standard cross-entropy loss, and vice versa. The CPS consistency has two roles: encourage high similarity between the predictions of two perturbed networks for the same input image, and expand training data by using the unlabeled data with pseudo labels. Experiment results show that our approach achieves the state-of-the-art semi-supervised segmentation performance on Cityscapes and PASCAL VOC 2012. Code is available at https://git.io/CPS.

Citations

• Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning
• Adversarial Learning for Semi-Supervised Semantic Segmentation
• Deep High-Resolution Representation Learning for Visual Recognition
• Structured Consistency Loss for semi-supervised semantic segmentation
• Self-Training With Noisy Student Improves ImageNet Classification

Cited by

• Toward Stable Semi-Supervised Remote Sensing Segmentation via Co-Guidance and Co-Fusion
• ESRVS: Extreme Semi-Supervised Retinal Vessel Segmentation with a Single Annotated Image
• Bi-Level Collaborative Learning for Few-Shot Scribble-Supervised Medical Image Segmentation
• Endo-SemiS: Towards Robust Semi-Supervised Image Segmentation for Endoscopic Video
• BiCoRec: Bias-Mitigated Context-Aware Sequential Recommendation Model
• Contrastive Learning for Semi-Supervised Deep Regression with Generalized Ordinal Rankings from Spectral Seriation
• Towards Robust Pseudo-Label Learning in Semantic Segmentation: An Encoding Perspective
• Vision-Language Enhanced Foundation Model for Semi-supervised Medical Image Segmentation
• Hierarchical Semi-Supervised Active Learning for Remote Sensing
• CoTBox-TTT: Grounding Medical VQA with Visual Chain-of-Thought Boxes During Test-time Training
• REASON: Probability map-guided dual-branch fusion framework for gastric content assessment
• Cross-pyramid consistency regularization for semi-supervised medical image segmentation
• DuetMatch: Harmonizing Semi-Supervised Brain MRI Segmentation via Decoupled Branch Optimization
• SEAL: Semantic-Aware Hierarchical Learning for Generalized Category Discovery
• BARL: Bilateral Alignment in Representation and Label Spaces for Semi-Supervised Volumetric Medical Image Segmentation
• Instance-Aware Robust Consistency Regularization for Semi-Supervised Nuclei Instance Segmentation
• Label-Efficient Cross-Modality Generalization for Liver Segmentation in Multi-Phase MRI
• Dual-supervised Asymmetric Co-training for Semi-supervised Medical Domain Generalization
• SCOUT: Semi-supervised Camouflaged Object Detection by Utilizing Text and Adaptive Data Selection
• When Confidence Fails: Revisiting Pseudo-Label Selection in Semi-supervised Semantic Segmentation
• Semi-MoE: Mixture-of-Experts meets Semi-Supervised Histopathology Segmentation
• Enhancing Dual Network Based Semi-Supervised Medical Image Segmentation with Uncertainty-Guided Pseudo-Labeling
• Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning
• Leveraging Information Divergence for Robust Semi-Supervised Fetal Ultrasound Image Segmentation
• MetaSSL: A General Heterogeneous Loss for Semi-Supervised Medical Image Segmentation
• Diversity-enhanced Collaborative Mamba for Semi-supervised Medical Image Segmentation
• SingleStrip: learning skull-stripping from a single labeled example
• Diverse Teaching and Label Propagation for Generic Semi-Supervised Medical Image Segmentation
• Uncertainty-aware Cross-training for Semi-supervised Medical Image Segmentation
• Iterative pseudo-labeling based adaptive copy-paste supervision for semi-supervised tumor segmentation
• M³HL: Mutual Mask Mix with High-Low Level Feature Consistency for Semi-Supervised Medical Image Segmentation
• Semi-Supervised Dual-Threshold Contrastive Learning for Ultrasound Image Classification and Segmentation
• ConformalSAM: Unlocking the Potential of Foundational Segmentation Models in Semi-Supervised Semantic Segmentation with Conformal Prediction
• Frequency Prior Guided Matching: A Data Augmentation Approach for Generalizable Semi-Supervised Polyp Segmentation
• Mitigating Spurious Correlations in Weakly Supervised Semantic Segmentation via Cross-architecture Consistency Regularization
• Dual Cross-image Semantic Consistency with Self-aware Pseudo Labeling for Semi-supervised Medical Image Segmentation
• Ensemble Foreground Management for Unsupervised Object Discovery
• Extreme Cardiac MRI Analysis under Respiratory Motion: Results of the CMRxMotion Challenge
• SemiSegECG: A Multi-Dataset Benchmark for Semi-Supervised Semantic Segmentation in ECG Delineation
• Comparison of Segmentation Methods in Remote Sensing for Land Use Land Cover
• RS-MTDF: Multi-Teacher Distillation and Fusion for Remote Sensing Semi-Supervised Semantic Segmentation
• Know Your Attention Maps: Class-specific Token Masking for Weakly Supervised Semantic Segmentation
• Learning Disentangled Stain and Structural Representations for Semi-Supervised Histopathology Segmentation
• Leveraging Out-of-Distribution Unlabeled Images: Semi-Supervised Semantic Segmentation with an Open-Vocabulary Model
• SSS: Semi-Supervised SAM-2 with Efficient Prompting for Medical Imaging Segmentation
• AdvMIM: Adversarial Masked Image Modeling for Semi-Supervised Medical Image Segmentation
• Baltimore Atlas: FreqWeaver Adapter for Semi-supervised Ultra-high Spatial Resolution Land Cover Classification
• Segmenting Visuals With Querying Words: Language Anchors For Semi-Supervised Image Segmentation
• Rethinking Semi-supervised Segmentation Beyond Accuracy: Reliability and Robustness
• Decoupled Competitive Framework for Semi-supervised Medical Image Segmentation
• RAIL: Region-Aware Instructive Learning for Semi-Supervised Tooth Segmentation in CBCT
• Adaptive Spatial Augmentation for Semi-supervised Semantic Segmentation
• CAST: Contrastive Adaptation and Distillation for Semi-Supervised Instance Segmentation
• Zero-Shot Pseudo Labels Generation Using SAM and CLIP for Semi-Supervised Semantic Segmentation
• Semi-Supervised Medical Image Segmentation via Dual Networks
• P3Net: Progressive and Periodic Perturbation for Semi-Supervised Medical Image Segmentation
• Mutual Evidential Deep Learning for Medical Image Segmentation
• Feedback-Driven Pseudo-Label Reliability Assessment: Redefining Thresholding for Semi-Supervised Semantic Segmentation
• Pseudo-supervised physics-informed neural networks with transfer learning for nonlinear bending analysis of metamaterial beams
• SuperCL: Superpixel Guided Contrastive Learning for Medical Image Segmentation Pre-training
• WT-BCP: Wavelet Transform based Bidirectional Copy-Paste for Semi-Supervised Medical Image Segmentation
• Cross-Frequency Collaborative Training Network and Dataset for Semi-supervised First Molar Root Canal Segmentation
• IGL-DT: Iterative Global-Local Feature Learning with Dual-Teacher Semantic Segmentation Framework under Limited Annotation Scheme
• HDC: Hierarchical Distillation for Multi-level Noisy Consistency in Semi-Supervised Fetal Ultrasound Segmentation
• Towards Micro-Action Recognition with Limited Annotations: An Asynchronous Pseudo Labeling and Training Approach
• SemiDAViL: Semi-supervised Domain Adaptation with Vision-Language Guidance for Semantic Segmentation

Related

• Semi-Supervised Semantic Segmentation with Cross-Consistency Training
• Revisiting Network Perturbation for Semi-Supervised Semantic Segmentation
• Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization
• GuidedMix-Net: Semi-supervised Semantic Segmentation by Using Labeled Images as Reference
• n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation