Points as Queries: Weakly Semi-supervised Object Detection by Points
Not on the map. Why?
Abstract (arXiv)
We propose a novel point annotated setting for the weakly semi-supervised object detection task, in which the dataset comprises small fully annotated images and large weakly annotated images by points. It achieves a balance between tremendous annotation burden and detection performance. Based on this setting, we analyze existing detectors and find that these detectors have difficulty in fully exploiting the power of the annotated points. To solve this, we introduce a new detector, Point DETR, which extends DETR by adding a point encoder. Extensive experiments conducted on MS-COCO dataset in various data settings show the effectiveness of our method. In particular, when using 20% fully labeled data from COCO, our detector achieves a promising performance, 33.3 AP, which outperforms a strong baseline (FCOS) by 2.0 AP, and we demonstrate the point annotations bring over 10 points in various AR metrics.
Key facts. Cited by 60 papers in the map (data through 7 Oct 2026). Builds on 28 of its 38 references that are in the map. First posted on arXiv 15 Apr 2021.
Builds on
28 of 38 references.
References · 28
- 2017Attention Is All You Needcited by 59,932
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Microsoft COCO: Common Objects in Contextcited by 17,498
- 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
- 2020End-to-End Object Detection with Transformerscited by 7,384
- 2016Feature Pyramid Networks for Object Detectioncited by 5,021
- 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,829
- 2019FCOS: Fully Convolutional One-Stage Object Detectioncited by 1,573
- 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,479
- 2016YOLO9000: Better, Faster, Strongercited by 1,429
- 2015Spatial Transformer Networkscited by 1,423
- 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,304
16 more
- 2019Objects as Pointscited by 1,172
- 2018Image Transformercited by 770
- 2018CornerNet: Detecting Objects as Paired Keypointscited by 763
- 2019CenterNet: Keypoint Triplets for Object Detectioncited by 654
- 2015What's the Point: Semantic Segmentation with Point Supervisioncited by 297
- 2020A Simple Semi-Supervised Learning Framework for Object Detectioncited by 282
- 2015Weakly Supervised Deep Detection Networkscited by 183
- 2020TIDE: A General Toolbox for Identifying Object Detection Errorscited by 113
- 2020Proposal Learning for Semi-Supervised Object Detectioncited by 55
- 2018Weakly Supervised Scene Parsing with Point-based Distance Metric Learningcited by 33
- 2018Towards Human-Machine Cooperation: Self-supervised Sample Mining for Object Detectioncited by 32
- 2017Deep Self-Taught Learning for Weakly Supervised Object Localizationcited by 27
- 2020UFO2: A Unified Framework towards Omni-supervised Object Detectioncited by 26
- 2017Soft Proposal Networks for Weakly Supervised Object Localizationcited by 21
- 2017Weakly Supervised Object Localization Using Things and Stuff Transfercited by 12
- 2017Weakly- and Semi-Supervised Object Detection with Expectation-Maximization Algorithmcited by 10
Built on by
Citing papers per year: 2021: 2; 2022: 18; 2023: 14; 2024: 15; 2025: 8; 2026: 3.
60 papers.
Sort:
- 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 613
- 2022Few-shot Object Counting and Detectioncited by 55
- 2023Mapping Degeneration Meets Label Evolution: Learning Infrared Small Target Detection with Single Point Supervisioncited by 50
- 2021Instance-Conditional Knowledge Distillation for Object Detectioncited by 42
- 2022Omni-DETR: Omni-Supervised Object Detection with Transformerscited by 28
- 2023PointOBB: Learning Oriented Object Detection via Single Point Supervisioncited by 25
- 2022Group R-CNN for Weakly Semi-supervised Object Detection with Pointscited by 23
- 2023The Devil is in the Points: Weakly Semi-Supervised Instance Segmentation via Point-Guided Mask Representationcited by 21
- 2023Point2RBox: Combine Knowledge from Synthetic Visual Patterns for End-to-end Oriented Object Detection with Single Point Supervisioncited by 21
- 2022Pointly-Supervised Panoptic Segmentationcited by 20
- 2023Monte Carlo Linear Clustering with Single-Point Supervision is Enough for Infrared Small Target Detectioncited by 19
- 2022Active Learning Strategies for Weakly-supervised Object Detectioncited by 17
38 more
- 2023Shepherding Slots to Objects: Towards Stable and Robust Object-Centric Learningcited by 17
- 2022Active Pointly-Supervised Instance Segmentationcited by 15
- 2022Interactive Multi-Class Tiny-Object Detectioncited by 14
- 2024Frontiers in Intelligent Colonoscopycited by 13
- 2022Point-Teaching: Weakly Semi-Supervised Object Detection with Point Annotationscited by 10
- 2022Visual Recognition by Requestcited by 10
- 2024From Easy to Hard: Progressive Active Learning Framework for Infrared Small Target Detection with Single Point Supervisioncited by 10
- 2022SAP-DETR: Bridging the Gap Between Salient Points and Queries-Based Transformer Detector for Fast Model Convergencycited by 9
- 2025PointOBB-v3: Expanding Performance Boundaries of Single Point-Supervised Oriented Object Detectioncited by 9
- 2022Graph Fusion Network for Multi-Oriented Object Detectioncited by 7
- 2022Scaling Novel Object Detection with Weakly Supervised Detection Transformerscited by 7
- 2025Wholly-WOOD: Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detectioncited by 6
- 2023Semi-supervised Object Detection: A Survey on Recent Research and Progresscited by 5
- 2024What is Point Supervision Worth in Video Instance Segmentation?cited by 5
- 2023Meta-tuning Loss Functions and Data Augmentation for Few-shot Object Detectioncited by 4
- 2024Semi-Supervised Object Detection: A Survey on Progress from CNN to Transformercited by 4
- 2024POLO -- Point-based, multi-class animal detectioncited by 4
- 2025LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectivescited by 4
- 2024Point-DETR3D: Leveraging Imagery Data with Spatial Point Prior for Weakly Semi-supervised 3D Object Detectioncited by 3
- 2024Towards Label-Efficient Human Matting: A Simple Baseline for Weakly Semi-Supervised Trimap-Free Human Mattingcited by 3
- 2025From Sight to Insight: Unleashing Eye-Tracking in Weakly Supervised Video Salient Object Detectioncited by 3
- 2025Partial Weakly-Supervised Oriented Object Detectioncited by 3
- 2025OmViD: Omni-supervised active learning for video action detectioncited by 3
- 2021Frequency Spectrum Augmentation Consistency for Domain Adaptive Object Detectioncited by 2
- 2022Semi-Weakly Supervised Object Detection by Sampling Pseudo Ground-Truth Boxescited by 2
- 2024On-the-Fly Point Annotation for Fast Medical Video Labelingcited by 2
- 2024Boosting Medical Image-based Cancer Detection via Text-guided Supervision from Reportscited by 2
- 2024A Point-Neighborhood Learning Framework for Nasal Endoscope Image Segmentationcited by 2
- 2022Bounding Box Priors for Cell Detection with Point Annotationscited by 1
- 2023Towards Universal Vision-language Omni-supervised Segmentationcited by 1
- 2023Deep Omni-supervised Learning for Rib Fracture Detection from Chest Radiology Imagescited by 1
- 2023Weakly-semi-supervised object detection in remotely sensed imagerycited by 1
- 2024Sparse Generation: Making Pseudo Labels Sparse for Point Weakly Supervised Object Detection on Low Data Volumecited by 1
- 2024Detecting Endangered Marine Species in Autonomous Underwater Vehicle Imagery Using Point Annotations and Few-Shot Learningcited by 1
- 2024UAVDB: Point-Guided Masks for UAV Detection and Segmentationcited by 1
- 2025P2Object: Single Point Supervised Object Detection and Instance Segmentationcited by 1
- 2026DExTeR: Weakly Semi-Supervised Object Detection with Class and Instance Experts for Medical Imagingcited by 1
- 2026OnPoint: Offline-to-Online Multi-Level Distillation for Point-Supervised Online Temporal Action Localizationcited by 1
50 most cited of 60 All 60 (JSON)
- 2026OnPoint: Offline-to-Online Multi-Level Distillation for Point-Supervised Online Temporal Action Localization1 Jul 2026
- 2026SPWOOD: Sparse Partial Weakly-Supervised Oriented Object Detection3 Feb 2026
- 2026DExTeR: Weakly Semi-Supervised Object Detection with Class and Instance Experts for Medical Imaging20 Jan 2026
- 2025ABBSPO: Adaptive Bounding Box Scaling and Symmetric Prior based Orientation Prediction for Detecting Aerial Image Objects10 Dec 2025
- 2025OmViD: Omni-supervised active learning for video action detection19 Aug 2025
- 2025Partial Weakly-Supervised Oriented Object Detection3 Jul 2025
- 2025From Sight to Insight: Unleashing Eye-Tracking in Weakly Supervised Video Salient Object Detection30 Jun 2025
- 2025P2Object: Single Point Supervised Object Detection and Instance Segmentation10 Apr 2025
- 2025LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectives24 Mar 2025
- 2025Wholly-WOOD: Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detection13 Feb 2025
- 2025PointOBB-v3: Expanding Performance Boundaries of Single Point-Supervised Oriented Object Detection23 Jan 2025
- 2024BCR-Net: Boundary-Category Refinement Network for Weakly Semi-Supervised X-Ray Prohibited Item Detection with Points25 Dec 2024
38 more
- 2024From Easy to Hard: Progressive Active Learning Framework for Infrared Small Target Detection with Single Point Supervision15 Dec 2024
- 2024Frontiers in Intelligent Colonoscopy22 Oct 2024
- 2024POLO -- Point-based, multi-class animal detection15 Oct 2024
- 2024UAVDB: Point-Guided Masks for UAV Detection and Segmentation9 Sep 2024
- 2024Semi-Supervised Object Detection: A Survey on Progress from CNN to Transformer11 Jul 2024
- 2024HUWSOD: Holistic Self-training for Unified Weakly Supervised Object Detection27 Jun 2024
- 2024Detecting Endangered Marine Species in Autonomous Underwater Vehicle Imagery Using Point Annotations and Few-Shot Learning4 Jun 2024
- 2024A Point-Neighborhood Learning Framework for Nasal Endoscope Image Segmentation30 May 2024
- 2024Boosting Medical Image-based Cancer Detection via Text-guided Supervision from Reports23 May 2024
- 2024On-the-Fly Point Annotation for Fast Medical Video Labeling22 Apr 2024
- 2024What is Point Supervision Worth in Video Instance Segmentation?1 Apr 2024
- 2024Towards Label-Efficient Human Matting: A Simple Baseline for Weakly Semi-Supervised Trimap-Free Human Matting1 Apr 2024
- 2024Sparse Generation: Making Pseudo Labels Sparse for Point Weakly Supervised Object Detection on Low Data Volume28 Mar 2024
- 2024Point-DETR3D: Leveraging Imagery Data with Spatial Point Prior for Weakly Semi-supervised 3D Object Detection22 Mar 2024
- 2023Weakly-semi-supervised object detection in remotely sensed imagery29 Nov 2023
- 2023Point2RBox: Combine Knowledge from Synthetic Visual Patterns for End-to-end Oriented Object Detection with Single Point Supervision23 Nov 2023
- 2023PointOBB: Learning Oriented Object Detection via Single Point Supervision23 Nov 2023
- 2023Click on Mask: A Labor-efficient Annotation Framework with Level Set for Infrared Small Target Detection19 Oct 2023
- 2023Semi-supervised Object Detection: A Survey on Recent Research and Progress25 Jun 2023
- 2023Deep Omni-supervised Learning for Rib Fracture Detection from Chest Radiology Images23 Jun 2023
- 2023Towards Precise Weakly Supervised Object Detection via Interactive Contrastive Learning of Context Information27 Apr 2023
- 2023Meta-tuning Loss Functions and Data Augmentation for Few-shot Object Detection24 Apr 2023
- 2023Monte Carlo Linear Clustering with Single-Point Supervision is Enough for Infrared Small Target Detection10 Apr 2023
- 2023Mapping Degeneration Meets Label Evolution: Learning Infrared Small Target Detection with Single Point Supervision4 Apr 2023
- 2023Shepherding Slots to Objects: Towards Stable and Robust Object-Centric Learning31 Mar 2023
- 2023The Devil is in the Points: Weakly Semi-Supervised Instance Segmentation via Point-Guided Mask Representation27 Mar 2023
- 2023Towards Universal Vision-language Omni-supervised Segmentation12 Mar 2023
- 2023OSRE: Object-to-Spot Rotation Estimation for Bike Parking Assessment1 Mar 2023
- 2022Bounding Box Priors for Cell Detection with Point Annotations11 Nov 2022
- 2022SAP-DETR: Bridging the Gap Between Salient Points and Queries-Based Transformer Detector for Fast Model Convergency3 Nov 2022
- 2022Pointly-Supervised Panoptic Segmentation25 Oct 2022
- 2022ComplETR: Reducing the cost of annotations for object detection in dense scenes with vision transformers13 Sep 2022
- 2022A Benchmark for Weakly Semi-Supervised Abnormality Localization in Chest X-Rays5 Sep 2022
- 2022Visual Recognition by Request28 Jul 2022
- 2022Active Learning Strategies for Weakly-supervised Object Detection25 Jul 2022
- 2022Active Pointly-Supervised Instance Segmentation23 Jul 2022
- 2022Few-shot Object Counting and Detection22 Jul 2022
- 2022Scaling Novel Object Detection with Weakly Supervised Detection Transformers11 Jul 2022
50 newest of 60 All 60 (JSON)
- 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 613
- 2022Few-shot Object Counting and Detectioncited by 55
- 2023Mapping Degeneration Meets Label Evolution: Learning Infrared Small Target Detection with Single Point Supervisioncited by 50
- 2021Instance-Conditional Knowledge Distillation for Object Detectioncited by 42
- 2022Omni-DETR: Omni-Supervised Object Detection with Transformerscited by 28
- 2023PointOBB: Learning Oriented Object Detection via Single Point Supervisioncited by 25
- 2022Group R-CNN for Weakly Semi-supervised Object Detection with Pointscited by 23
- 2023The Devil is in the Points: Weakly Semi-Supervised Instance Segmentation via Point-Guided Mask Representationcited by 21
- 2023Point2RBox: Combine Knowledge from Synthetic Visual Patterns for End-to-end Oriented Object Detection with Single Point Supervisioncited by 21
- 2022Pointly-Supervised Panoptic Segmentationcited by 20
- 2023Monte Carlo Linear Clustering with Single-Point Supervision is Enough for Infrared Small Target Detectioncited by 19
- 2022Active Learning Strategies for Weakly-supervised Object Detectioncited by 17
12 most cited of 60 All 60 (JSON)
- 2026OnPoint: Offline-to-Online Multi-Level Distillation for Point-Supervised Online Temporal Action Localization1 Jul 2026
- 2026SPWOOD: Sparse Partial Weakly-Supervised Oriented Object Detection3 Feb 2026
- 2026DExTeR: Weakly Semi-Supervised Object Detection with Class and Instance Experts for Medical Imaging20 Jan 2026
- 2025ABBSPO: Adaptive Bounding Box Scaling and Symmetric Prior based Orientation Prediction for Detecting Aerial Image Objects10 Dec 2025
- 2025OmViD: Omni-supervised active learning for video action detection19 Aug 2025
- 2025Partial Weakly-Supervised Oriented Object Detection3 Jul 2025
- 2025From Sight to Insight: Unleashing Eye-Tracking in Weakly Supervised Video Salient Object Detection30 Jun 2025
- 2025P2Object: Single Point Supervised Object Detection and Instance Segmentation10 Apr 2025
- 2025LiDAR Remote Sensing Meets Weak Supervision: Concepts, Methods, and Perspectives24 Mar 2025
- 2025Wholly-WOOD: Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detection13 Feb 2025
- 2025PointOBB-v3: Expanding Performance Boundaries of Single Point-Supervised Oriented Object Detection23 Jan 2025
- 2024BCR-Net: Boundary-Category Refinement Network for Weakly Semi-Supervised X-Ray Prohibited Item Detection with Points25 Dec 2024
12 newest of 60 All 60 (JSON)