mlsanityData to 7 Oct 2026

MC-PanDA: Mask Confidence for Panoptic Domain Adaptation

arXiv 2407.14110 ↗First posted 19 Jul 2024cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

Domain adaptive panoptic segmentation promises to resolve the long tail of corner cases in natural scene understanding. Previous state of the art addresses this problem with cross-task consistency, careful system-level optimization and heuristic improvement of teacher predictions. In contrast, we propose to build upon remarkable capability of mask transformers to estimate their own prediction uncertainty. Our method avoids noise amplification by leveraging fine-grained confidence of panoptic teacher predictions. In particular, we modulate the loss with mask-wide confidence and discourage back-propagation in pixels with uncertain teacher or confident student. Experimental evaluation on standard benchmarks reveals a substantial contribution of the proposed selection techniques. We report 47.4 PQ on Synthia to Cityscapes, which corresponds to an improvement of 6.2 percentage points over the state of the art. The source code is available at https://github.com/helen1c/MC-PanDA.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 51 of its 70 references that are in the map. First posted on arXiv 19 Jul 2024.

Builds on

51 of 70 references.

References · 51

  1. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  4. 2020End-to-End Object Detection with Transformerscited by 7,384
  5. 2017Mask R-CNNcited by 7,259
  6. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  7. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,410
  8. 2015SSD: Single Shot MultiBox Detectorcited by 3,125
  9. 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,923
  10. 2021Masked-attention Mask Transformer for Universal Image Segmentationcited by 2,338
  11. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  12. 2016Deep CORAL: Correlation Alignment for Deep Domain Adaptationcited by 1,194
38 more
  1. 2021Per-Pixel Classification is Not All You Need for Semantic Segmentationcited by 1,148
  2. 2017Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learningcited by 918
  3. 2016Playing for Data: Ground Truth from Computer Gamescited by 816
  4. 2018Panoptic Segmentationcited by 796
  5. 2021Domain Generalization: A Surveycited by 760
  6. 2019Panoptic Feature Pyramid Networkscited by 644
  7. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  8. 2019PointRend: Image Segmentation as Renderingcited by 439
  9. 2017Semantic Foggy Scene Understanding with Synthetic Datacited by 401
  10. 2019Panoptic-DeepLabcited by 362
  11. 2022Mask DINO: Towards A Unified Transformer-based Framework for Object Detection and Segmentationcited by 341
  12. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 299
  13. 2014Understanding image representations by measuring their equivariance and equivalencecited by 240
  14. 2020DACS: Domain Adaptation via Cross-domain Mixed Samplingcited by 226
  15. 2020Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentationcited by 222
  16. 2020ClassMix: Segmentation-Based Data Augmentation for Semi-Supervised Learningcited by 209
  17. 2022MIC: Masked Image Consistency for Context-Enhanced Domain Adaptationcited by 176
  18. 2019UPSNet: A Unified Panoptic Segmentation Networkcited by 175
  19. 2022HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentationcited by 163
  20. 2020Video Panoptic Segmentationcited by 141
  21. 2019The Fishyscapes Benchmark: Measuring Blind Spots in Semantic Segmentationcited by 122
  22. 2021MSeg: A Composite Dataset for Multi-domain Semantic Segmentationcited by 122
  23. 2021Context-Aware Mixup for Domain Adaptive Semantic Segmentationcited by 79
  24. 2022RbA: Segmenting Unknown Regions Rejected by Allcited by 67
  25. 2020Uncertainty-Aware Consistency Regularization for Cross-Domain Semantic Segmentationcited by 45
  26. 2023Unmasking Anomalies in Road-Scene Segmentationcited by 45
  27. 2023On Advantages of Mask-level Recognition for Outlier-aware Segmentationcited by 42
  28. 2021Cross-View Regularization for Domain Adaptive Panoptic Segmentationcited by 41
  29. 2022PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentationcited by 41
  30. 2023Maskomaly:Zero-Shot Mask Anomaly Segmentationcited by 34
  31. 2024Outlier detection by ensembling uncertainty with negative objectnesscited by 34
  32. 2018Learning to Fuse Things and Stuffcited by 26
  33. 2023Domain Adaptive and Generalizable Network Architectures and Training Strategies for Semantic Image Segmentationcited by 26
  34. 2023EDAPS: Enhanced Domain-Adaptive Panoptic Segmentationcited by 17
  35. 2023Guided Distillation for Semi-Supervised Instance Segmentationcited by 12
  36. 2022The Missing Link: Finding label relations across datasetscited by 9
  37. 2022UniDAformer: Unified Domain Adaptive Panoptic Segmentation Transformer via Hierarchical Mask Calibrationcited by 8
  38. 2022Weakly supervised training of universal visual concepts for multi-domain semantic segmentationcited by 5

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2025Unlocking Constraints: Source-Free Occlusion-Aware Seamless Segmentationcited by 3
  2. 2025DEARLi: Decoupled Enhancement of Recognition and Localization for Semi-supervised Panoptic Segmentationcited by 3
  3. 2026PanDA: Unsupervised Domain Adaptation for Multimodal 3D Panoptic Segmentation in Autonomous Drivingcited by 0
  4. 2026MC-PanDA++: Simpler, Stronger, and More Robust Domain-Adaptive Panoptic Segmentationcited by 0