Active Learning with Context Sampling and One-vs-Rest Entropy for Semantic Segmentation
Not on the map. Why?
Abstract (arXiv)
Multi-class semantic segmentation remains a cornerstone challenge in computer vision. Yet, dataset creation remains excessively demanding in time and effort, especially for specialized domains. Active Learning (AL) mitigates this challenge by selecting data points for annotation strategically. However, existing patch-based AL methods often overlook boundary pixels critical information, essential for accurate segmentation. We present OREAL, a novel patch-based AL method designed for multi-class semantic segmentation. OREAL enhances boundary detection by employing maximum aggregation of pixel-wise uncertainty scores. Additionally, we introduce one-vs-rest entropy, a novel uncertainty score function that computes class-wise uncertainties while achieving implicit class balancing during dataset creation. Comprehensive experiments across diverse datasets and model architectures validate our hypothesis.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 29 of its 39 references that are in the map. First posted on arXiv 9 Dec 2024.
Builds on
29 of 39 references.
References · 29
- 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
- 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,161
- 2021Segmenter: Transformer for Semantic Segmentationcited by 855
- 2015What's the Point: Semantic Segmentation with Point Supervisioncited by 298
- 2019Variational Adversarial Active Learningcited by 284
- 2017Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentationcited by 146
- 20202018 Robotic Scene Segmentation Challengecited by 124
- 2019ViewAL: Active Learning with Viewpoint Entropy for Semantic Segmentationcited by 96
- 2013SEEDS: Superpixels Extracted via Energy-Driven Samplingcited by 75
- 2021Towards Fewer Annotations: Active Learning via Region Impurity and Prediction Uncertainty for Domain Adaptive Semantic Segmentationcited by 57
- 2020Reinforced active learning for image segmentationcited by 56
- 2018CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentationcited by 49
17 more
- 2021DSAL: Deeply Supervised Active Learning from Strong and Weak Labelers for Biomedical Image Segmentationcited by 27
- 2020MetaBox+: A new Region Based Active Learning Method for Semantic Segmentation using Priority Mapscited by 23
- 2020Importance of Self-Consistency in Active Learning for Semantic Segmentationcited by 21
- 2020Suggestive Annotation of Brain Tumour Images with Gradient-guided Samplingcited by 19
- 2023Algorithm Selection for Deep Active Learning with Imbalanced Datasetscited by 18
- 2022DIAL: Deep Interactive and Active Learning for Semantic Segmentation in Remote Sensingcited by 17
- 2021Class-Balanced Active Learning for Image Classificationcited by 15
- 2023Adaptive Superpixel for Active Learning in Semantic Segmentationcited by 14
- 2022Minority Class Oriented Active Learning for Imbalanced Datasetscited by 11
- 2022CLINICAL: Targeted Active Learning for Imbalanced Medical Image Classificationcited by 9
- 2023Active Learning for Semantic Segmentation with Multi-class Label Querycited by 9
- 2020Active Class Incremental Learning for Imbalanced Datasetscited by 8
- 2021Reducing Annotating Load: Active Learning with Synthetic Images in Surgical Instrument Segmentationcited by 6
- 2021MEAL: Manifold Embedding-based Active Learningcited by 5
- 2021Active learning for imbalanced data under cold startcited by 5
- 2023Correlation-aware active learning for surgery video segmentationcited by 4
- 2021Weakly-Supervised Semantic Segmentation by Learning Label Uncertaintycited by 2
Built on by
Citing papers per year: 2025: 2.
2 papers.