mlsanityData to 14 Oct 2026

Mining of Single-Class by Active Learning for Semantic Segmentation

arXiv 2307.09109 ↗First posted 18 Jul 2023cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Several Active Learning (AL) policies require retraining a target model several times in order to identify the most informative samples and rarely offer the option to focus on the acquisition of samples from underrepresented classes. Here the Mining of Single-Class by Active Learning (MiSiCAL) paradigm is introduced where an AL policy is constructed through deep reinforcement learning and exploits quantity-accuracy correlations to build datasets on which high-performance models can be trained with regards to specific classes. MiSiCAL is especially helpful in the case of very large batch sizes since it does not require repeated model training sessions as is common in other AL methods. This is thanks to its ability to exploit fixed representations of the candidate data points. We find that MiSiCAL is able to outperform a random policy on 150 out of 171 COCO10k classes, while the strongest baseline only outperforms random on 101 classes.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 43 of its 65 references that are in the map. First posted on arXiv 18 Jul 2023.

Builds on

43 of 65 references.

References · 43

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  3. 2022Training language models to follow instructions with human feedbackcited by 15,971
  4. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  5. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
  6. 2015Continuous control with deep reinforcement learningcited by 3,407
  7. 2019Searching for MobileNetV3cited by 1,994
  8. 2015Deep Reinforcement Learning with Double Q-learningcited by 1,745
  9. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,616
  10. 2021TransUNet: Transformers Make Strong Encoders for Medical Image Segmentationcited by 1,558
  11. 2015Prioritized Experience Replaycited by 1,099
  12. 2016COCO-Stuff: Thing and Stuff Classes in Contextcited by 864
31 more
  1. 2017Deep Bayesian Active Learning with Image Datacited by 819
  2. 2017Rainbow: Combining Improvements in Deep Reinforcement Learningcited by 786
  3. 2019Data Shapley: Equitable Valuation of Data for Machine Learningcited by 629
  4. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  5. 2022Vision Transformer Adapter for Dense Predictionscited by 565
  6. 2020Scaling Laws for Autoregressive Generative Modelingcited by 522
  7. 2022Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Taskscited by 511
  8. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 432
  9. 2021Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image Analysiscited by 422
  10. 2021Medical Transformer: Gated Axial-Attention for Medical Image Segmentationcited by 320
  11. 2019Variational Adversarial Active Learningcited by 284
  12. 2020Revisiting Fundamentals of Experience Replaycited by 175
  13. 2021UTNet: A Hybrid Transformer Architecture for Medical Image Segmentationcited by 116
  14. 2019Discriminative Active Learningcited by 98
  15. 2021Influence Selection for Active Learningcited by 84
  16. 2017Learning how to Active Learn: A Deep Reinforcement Learning Approachcited by 79
  17. 2022Unpacking Reward Shaping: Understanding the Benefits of Reward Engineering on Sample Complexitycited by 62
  18. 2020Reinforced active learning for image segmentationcited by 56
  19. 2022A Survey of Active Learning for Natural Language Processingcited by 40
  20. 2019Parting with Illusions about Deep Active Learningcited by 38
  21. 2021Self-Supervised Online Reward Shaping in Sparse-Reward Environmentscited by 37
  22. 2021On the Importance of Effectively Adapting Pretrained Language Models for Active Learningcited by 32
  23. 2019Understanding Multi-Step Deep Reinforcement Learning: A Systematic Study of the DQN Targetcited by 31
  24. 2021All you need are a few pixels: semantic segmentation with PixelPickcited by 31
  25. 2021Data Shapley Valuation for Efficient Batch Active Learningcited by 27
  26. 2018A Brief Survey and an Application of Semantic Image Segmentation for Autonomous Drivingcited by 19
  27. 2020The Effect of Multi-step Methods on Overestimation in Deep Reinforcement Learningcited by 14
  28. 2020On uncertainty estimation in active learning for image segmentationcited by 9
  29. 2022Is More Data Better? Re-thinking the Importance of Efficiency in Abusive Language Detection with Transformers-Based Active Learningcited by 4
  30. 2021Feature Selection for Imbalanced Data with Deep Sparse Autoencoders Ensemblecited by 3
  31. 2022Deep reinforced active learning for multi-class image classificationcited by 3

Built on by

Not cited yet.