mlsanityData to 7 Oct 2026

Training Region-based Object Detectors with Online Hard Example Mining

arXiv 1604.03540 ↗First posted 12 Apr 2016cs.CV cs.LGcited by 577

Not on the map. Why?

Abstract (arXiv)

The field of object detection has made significant advances riding on the wave of region-based ConvNets, but their training procedure still includes many heuristics and hyperparameters that are costly to tune. We present a simple yet surprisingly effective online hard example mining (OHEM) algorithm for training region-based ConvNet detectors. Our motivation is the same as it has always been -- detection datasets contain an overwhelming number of easy examples and a small number of hard examples. Automatic selection of these hard examples can make training more effective and efficient. OHEM is a simple and intuitive algorithm that eliminates several heuristics and hyperparameters in common use. But more importantly, it yields consistent and significant boosts in detection performance on benchmarks like PASCAL VOC 2007 and 2012. Its effectiveness increases as datasets become larger and more difficult, as demonstrated by the results on the MS COCO dataset. Moreover, combined with complementary advances in the field, OHEM leads to state-of-the-art results of 78.9% and 76.3% mAP on PASCAL VOC 2007 and 2012 respectively.

Key facts. Cited by 577 papers in the map (data through 7 Oct 2026). First posted on arXiv 12 Apr 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 137; 2022: 119; 2023: 105; 2024: 82; 2025: 49; 2026: 85.

577 papers.

Sort:

By area:

  1. 2022AdaFace: Quality Adaptive Margin for Face Recognitioncited by 390
  2. 2023Occ3D: A Large-Scale 3D Occupancy Prediction Benchmark for Autonomous Drivingcited by 325
  3. 2021End-to-End Semi-Supervised Object Detection with Soft Teachercited by 315
  4. 2021Unbiased Teacher for Semi-Supervised Object Detectioncited by 305
  5. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 299
  6. 2021Curriculum Learning: A Surveycited by 257
  7. 2024VideoLLM-online: Online Video Large Language Model for Streaming Videocited by 216
  8. 2021Localizing Visual Sounds the Hard Waycited by 169
  9. 2021FSCE: Few-Shot Object Detection via Contrastive Proposal Encodingcited by 168
  10. 2023EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latenciescited by 157
  11. 2021Delving into Localization Errors for Monocular 3D Object Detectioncited by 138
  12. 2021General Instance Distillation for Object Detectioncited by 113
38 more
  1. 2021LaneRCNN: Distributed Representations for Graph-Centric Motion Forecastingcited by 112
  2. 2024Improving the Training of Rectified Flowscited by 106
  3. 2021Semi-supervised Left Atrium Segmentation with Mutual Consistency Trainingcited by 104
  4. 2022Improving Generalization in Federated Learning by Seeking Flat Minimacited by 98
  5. 2021A Survey of Modern Deep Learning based Object Detection Modelscited by 97
  6. 2021Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenescited by 89
  7. 2022PIDNet: A Real-time Semantic Segmentation Network Inspired by PID Controllerscited by 84
  8. 2021Fourier Contour Embedding for Arbitrary-Shaped Text Detectioncited by 80
  9. 2022TransVOD: End-to-End Video Object Detection with Spatial-Temporal Transformerscited by 78
  10. 2021FASA: Feature Augmentation and Sampling Adaptation for Long-Tailed Instance Segmentationcited by 76
  11. 2021ArtiBoost: Boosting Articulated 3D Hand-Object Pose Estimation via Online Exploration and Synthesiscited by 76
  12. 2022A Survey on Long-Tailed Visual Recognitioncited by 75
  13. 2024A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and Localizationcited by 74
  14. 2021Context Decoupling Augmentation for Weakly Supervised Semantic Segmentationcited by 69
  15. 2021Learning Fast Sample Re-weighting Without Reward Datacited by 69
  16. 2022R2L: Distilling Neural Radiance Field to Neural Light Field for Efficient Novel View Synthesiscited by 69
  17. 2022Unbiased Teacher v2: Semi-supervised Object Detection for Anchor-free and Anchor-based Detectorscited by 69
  18. 2021Ground-aware Monocular 3D Object Detection for Autonomous Drivingcited by 65
  19. 2021PAN++: Towards Efficient and Accurate End-to-End Spotting of Arbitrarily-Shaped Textcited by 65
  20. 2022PolyLoss: A Polynomial Expansion Perspective of Classification Loss Functionscited by 62
  21. 2023FreeMask: Synthetic Images with Dense Annotations Make Stronger Segmentation Modelscited by 61
  22. 2023Multiple Instance Learning Framework with Masked Hard Instance Mining for Whole Slide Image Classificationcited by 58
  23. 2023FocalFormer3D : Focusing on Hard Instance for 3D Object Detectioncited by 58
  24. 2021Focal and Efficient IOU Loss for Accurate Bounding Box Regressioncited by 57
  25. 2022Detecting tiny objects in aerial images: A normalized Wasserstein distance and a new benchmarkcited by 56
  26. 2023Hard Patches Mining for Masked Image Modelingcited by 56
  27. 2024UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentationcited by 53
  28. 2023EarthVQA: Towards Queryable Earth via Relational Reasoning-Based Remote Sensing Visual Question Answeringcited by 51
  29. 2021On Exposing the Challenging Long Tail in Future Prediction of Traffic Actorscited by 48
  30. 2021Data-Uncertainty Guided Multi-Phase Learning for Semi-Supervised Object Detectioncited by 48
  31. 2021Hard Sample Aware Noise Robust Learning for Histopathology Image Classificationcited by 48
  32. 2021Adaptive Boundary Proposal Network for Arbitrary Shape Text Detectioncited by 47
  33. 2022Equalized Focal Loss for Dense Long-Tailed Object Detectioncited by 46
  34. 2022Exploring Structure-aware Transformer over Interaction Proposals for Human-Object Interaction Detectioncited by 46
  35. 2022Behind Every Domain There is a Shift: Adapting Distortion-aware Vision Transformers for Panoramic Semantic Segmentationcited by 46
  36. 2021Instance-wise Hard Negative Example Generation for Contrastive Learning in Unpaired Image-to-Image Translationcited by 45
  37. 2023CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentationcited by 43
  38. 2024Not All Voxels Are Equal: Hardness-Aware Semantic Scene Completion with Self-Distillationcited by 43

50 most cited of 577 All 577 (JSON)