mlsanityData to 7 Oct 2026

Rethinking Post-Hoc Calibration in Semantic Segmentation

arXiv 2607.01902 ↗First posted 2 Jul 2026cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Reliable confidence estimates are essential in semantic segmentation, yet modern models often remain miscalibrated. We investigate two overlooked issues in post-hoc calibration. First, adding a constant to all logits leaves softmax probabilities unchanged, but several standard calibrators depend on this arbitrary offset. In segmentation, this offset can vary across pixels or voxels, introducing spatially varying representation dependence. We characterize translation-invariant (TI) calibrators and construct TI counterparts of shift-sensitive methods. Second, calibrating with cross-entropy can degrade segmentation quality due to mismatched training and calibration objectives and limited calibration data. We investigate decision-preserving calibration under argmax- and order-preservation constraints. Since these constraints restrict affine softmax calibrators to temperature scaling, we introduce more expressive class-conditional affine calibrators that preserve decisions. Across natural-image and medical segmentation benchmarks, including corruption-based covariate shift, TI variants generally improve calibration, while decision-preserving variants prevent segmentation degradation by construction and retain strong calibration performance. Our findings provide practical design principles for post-hoc calibration in semantic segmentation.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 29 of its 41 references that are in the map. First posted on arXiv 2 Jul 2026.

Builds on

29 of 41 references.

References · 29

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2017On Calibration of Modern Neural Networkscited by 4,860
  3. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  4. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,888
  5. 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,923
  6. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  7. 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,455
  8. 2020Energy-based Out-of-distribution Detectioncited by 1,296
  9. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,274
  10. 2019When Does Label Smoothing Help?cited by 985
  11. 2017Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentationscited by 614
  12. 2019Deep Ensembles: A Loss Landscape Perspectivecited by 438
17 more
  1. 2019Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibrationcited by 359
  2. 2020Calibrating Deep Neural Networks using Focal Losscited by 353
  3. 2020Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learningcited by 190
  4. 2019Confidence Calibration and Predictive Uncertainty Estimation for Deep Medical Image Segmentationcited by 162
  5. 2020Uncertainty Quantification and Deep Ensemblescited by 138
  6. 2019Benchmarking the Robustness of Semantic Segmentation Modelscited by 111
  7. 2020Local Temperature Scaling for Probability Calibrationcited by 95
  8. 2020Intra Order-preserving Functions for Calibration of Multi-Class Neural Networkscited by 65
  9. 2024The 2024 Brain Tumor Segmentation (BraTS) Challenge: Glioma Segmentation on Post-treatment MRIcited by 65
  10. 2019Non-Parametric Calibration for Classificationcited by 57
  11. 2018Dirichlet-based Gaussian Processes for Large-scale Calibrated Classificationcited by 50
  12. 2022On Calibrating Semantic Segmentation Models: Analyses and An Algorithmcited by 25
  13. 2024ValUES: A Framework for Systematic Validation of Uncertainty Estimation in Semantic Segmentationcited by 25
  14. 2024Comparative Benchmarking of Failure Detection Methods in Medical Image Segmentation: Unveiling the Role of Confidence Aggregationcited by 12
  15. 2025Spatially-Aware Evaluation of Segmentation Uncertaintycited by 6
  16. 2024Uncertainty Calibration with Energy Based Instance-wise Scaling in the Wild Datasetcited by 4
  17. 2025We Care Each Pixel: Calibrating on Medical Segmentation Modelcited by 2

Built on by

Not cited yet.