DeLR: Active Learning for Detection with Decoupled Localization and Recognition Query
Not on the map. Why?
Abstract (arXiv)
Active learning has been demonstrated effective to reduce labeling cost, while most progress has been designed for image recognition, there still lacks instance-level active learning for object detection. In this paper, we rethink two key components, i.e., localization and recognition, for object detection, and find that the correctness of them are highly related, therefore, it is not necessary to annotate both boxes and classes if we are given pseudo annotations provided with the trained model. Motivated by this, we propose an efficient query strategy, termed as DeLR, that Decoupling the Localization and Recognition for active query. In this way, we are probably free of class annotations when the localization is correct, and able to assign the labeling budget for more informative samples. There are two main differences in DeLR: 1) Unlike previous methods mostly focus on image-level annotations, where the queried samples are selected and exhausted annotated. In DeLR, the query is based on region-level, and we only annotate the object region that is queried; 2) Instead of directly providing both localization and recognition annotations, we separately query the two components, and thus reduce the recognition budget with the pseudo class labels provided by the model. Experiments on several benchmarks demonstrate its superiority. We hope our proposed query strategy would shed light on researches in active learning in object detection.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 28 of its 43 references that are in the map. First posted on arXiv 28 Dec 2023.
Builds on
28 of 43 references.
References · 28
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2014Microsoft COCO: Common Objects in Contextcited by 17,608
- 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
- 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
- 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
- 2016Feature Pyramid Networks for Object Detectioncited by 5,036
- 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
- 2019Learning Loss for Active Learningcited by 336
- 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 316
- 2015Weakly Supervised Deep Detection Networkscited by 183
- 2017Cost-Effective Active Learning for Deep Image Classificationcited by 166
- 2020Contextual Diversity for Active Learningcited by 139
16 more
- 2021Humble Teachers Teach Better Students for Semi-Supervised Object Detectioncited by 116
- 2019Consistency-based Semi-supervised Active Learning: Towards Minimizing Labeling Costcited by 114
- 2017Extreme clicking for efficient object annotationcited by 81
- 2021Active Learning for Deep Object Detection via Probabilistic Modelingcited by 79
- 2021Multiple instance active learning for object detectioncited by 77
- 2019C-MIL: Continuation Multiple Instance Learning for Weakly Supervised Object Detectioncited by 71
- 2019Active Learning for Deep Detection Neural Networkscited by 62
- 2021Semi-Supervised Active Learning with Temporal Output Discrepancycited by 50
- 2018Localization-Aware Active Learning for Object Detectioncited by 48
- 2019Min-Entropy Latent Model for Weakly Supervised Object Detectioncited by 45
- 2022Omni-DETR: Omni-Supervised Object Detection with Transformerscited by 29
- 2020UFO2: A Unified Framework towards Omni-supervised Object Detectioncited by 26
- 2015Hierarchical Subquery Evaluation for Active Learning on a Graphcited by 21
- 2016We don't need no bounding-boxes: Training object class detectors using only human verificationcited by 20
- 2019An Adaptive Supervision Framework for Active Learning in Object Detectioncited by 20
- 2020One-bit Supervision for Image Classificationcited by 10
Built on by
Citing papers per year: 2026: 2.
2 papers.