Data to 14 Oct 2026

ClassMix: Segmentation-Based Data Augmentation for Semi-Supervised Learning

arXiv 2007.07936 ↗First posted 15 Jul 2020cs.CVcited by 209

Not on the map. Why?

Abstract (arXiv)

The state of the art in semantic segmentation is steadily increasing in performance, resulting in more precise and reliable segmentations in many different applications. However, progress is limited by the cost of generating labels for training, which sometimes requires hours of manual labor for a single image. Because of this, semi-supervised methods have been applied to this task, with varying degrees of success. A key challenge is that common augmentations used in semi-supervised classification are less effective for semantic segmentation. We propose a novel data augmentation mechanism called ClassMix, which generates augmentations by mixing unlabelled samples, by leveraging on the network's predictions for respecting object boundaries. We evaluate this augmentation technique on two common semi-supervised semantic segmentation benchmarks, showing that it attains state-of-the-art results. Lastly, we also provide extensive ablation studies comparing different design decisions and training regimes.

Key facts. Cited by 209 papers in the map (data through 14 Oct 2026). First posted on arXiv 15 Jul 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 35; 2022: 50; 2023: 51; 2024: 36; 2025: 24; 2026: 13.

209 papers.

Sort:

By area:

2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 300
2022Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labelscited by 218
2021ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentationcited by 184
2021Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalizationcited by 141
2021Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bankcited by 125
2021Pixel Contrastive-Consistent Semi-Supervised Semantic Segmentationcited by 101
2022LaserMix for Semi-Supervised LiDAR Semantic Segmentationcited by 93
2022SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentationcited by 87
2022Self-Training: A Surveycited by 86
2021Bootstrapping Semantic Segmentation with Regional Contrastcited by 80
2021Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimationcited by 80
2021A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentationcited by 79

38 more

2021Context-Aware Mixup for Domain Adaptive Semantic Segmentationcited by 79
2021Re-distributing Biased Pseudo Labels for Semi-supervised Semantic Segmentation: A Baseline Investigationcited by 78
2022Semi-supervised Semantic Segmentation with Error Localization Networkcited by 57
2024Multi-Modal Data-Efficient 3D Scene Understanding for Autonomous Drivingcited by 54
2022Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularizationcited by 50
2022Rethinking Data Augmentation for Single-source Domain Generalization in Medical Image Segmentationcited by 45
2021Learning Multiple Dense Prediction Tasks from Partially Annotated Datacited by 42
2021ConDA: Unsupervised Domain Adaptation for LiDAR Segmentation via Regularized Domain Concatenationcited by 41
2022PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentationcited by 41
2023SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidancecited by 41
2023All for One, and One for All: UrbanSyn Dataset, the third Musketeer of Synthetic Driving Scenescited by 41
2021Every Annotation Counts: Multi-label Deep Supervision for Medical Image Segmentationcited by 37
2023Conflict-Based Cross-View Consistency for Semi-Supervised Semantic Segmentationcited by 35
2023One-shot Unsupervised Domain Adaptation with Personalized Diffusion Modelscited by 35
2023Semi-Supervised Semantic Segmentation via Gentle Teaching Assistantcited by 34
2021MixMo: Mixing Multiple Inputs for Multiple Outputs via Deep Subnetworkscited by 32
2021Data-Free Knowledge Transfer: A Surveycited by 31
2022Deliberated Domain Bridging for Domain Adaptive Semantic Segmentationcited by 31
2022A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainabilitycited by 31
2023CMDA: Cross-Modality Domain Adaptation for Nighttime Semantic Segmentationcited by 30
2021Self-Adversarial Disentangling for Specific Domain Adaptationcited by 29
2023Survey on Unsupervised Domain Adaptation for Semantic Segmentation for Visual Perception in Automated Drivingcited by 29
2022ConfMix: Unsupervised Domain Adaptation for Object Detection via Confidence-based Mixingcited by 28
2023Semantic Image Segmentation: Two Decades of Researchcited by 27
2023Domain Adaptive and Generalizable Network Architectures and Training Strategies for Semantic Image Segmentationcited by 26
2023Balancing Logit Variation for Long-tailed Semantic Segmentationcited by 26
2021Improving Semi-Supervised and Domain-Adaptive Semantic Segmentation with Self-Supervised Depth Estimationcited by 25
2023Image Data Augmentation Approaches: A Comprehensive Survey and Future directionscited by 23
2023Switching Temporary Teachers for Semi-Supervised Semantic Segmentationcited by 23
2022Decoder Denoising Pretraining for Semantic Segmentationcited by 20
2023Compositional Semantic Mix for Domain Adaptation in Point Cloud Segmentationcited by 20
2024A Survey on Mixup Augmentations and Beyondcited by 20
2021Robust Semantic Segmentation with Superpixel-Mixcited by 19
2023DiGA: Distil to Generalize and then Adapt for Domain Adaptive Semantic Segmentationcited by 19
2022Domain-Agnostic Prior for Transfer Semantic Segmentationcited by 18
2023CoDEPS: Online Continual Learning for Depth Estimation and Panoptic Segmentationcited by 18
2023A Survey on Semi-Supervised Semantic Segmentationcited by 17
2023EDAPS: Enhanced Domain-Adaptive Panoptic Segmentationcited by 17

50 most cited of 209 All 209 (JSON)