mlsanityData to 7 Oct 2026

Balanced Meta-Softmax for Long-Tailed Visual Recognition

arXiv 2007.10740 ↗First posted 21 Jul 2020cs.LG cs.CV stat.MLcited by 397

Not on the map. Why?

Abstract (arXiv)

Deep classifiers have achieved great success in visual recognition. However, real-world data is long-tailed by nature, leading to the mismatch between training and testing distributions. In this paper, we show that the Softmax function, though used in most classification tasks, gives a biased gradient estimation under the long-tailed setup. This paper presents Balanced Softmax, an elegant unbiased extension of Softmax, to accommodate the label distribution shift between training and testing. Theoretically, we derive the generalization bound for multiclass Softmax regression and show our loss minimizes the bound. In addition, we introduce Balanced Meta-Softmax, applying a complementary Meta Sampler to estimate the optimal class sample rate and further improve long-tailed learning. In our experiments, we demonstrate that Balanced Meta-Softmax outperforms state-of-the-art long-tailed classification solutions on both visual recognition and instance segmentation tasks.

Key facts. Cited by 397 papers in the map (data through 7 Oct 2026). First posted on arXiv 21 Jul 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 38; 2022: 60; 2023: 93; 2024: 94; 2025: 53; 2026: 59.

397 papers.

Sort:

By area:

  1. 2021Deep Long-Tailed Learning: A Surveycited by 416
  2. 2021Parametric Contrastive Learningcited by 217
  3. 2023Towards Open Vocabulary Learning: A Surveycited by 165
  4. 2021Reviving Iterative Training with Mask Guidance for Interactive Segmentationcited by 162
  5. 2022Balanced Contrastive Learning for Long-Tailed Visual Recognitioncited by 143
  6. 2023Change is Hard: A Closer Look at Subpopulation Shiftcited by 119
  7. 2022Parametric Classification for Generalized Category Discovery: A Baseline Studycited by 113
  8. 2022Long-Tailed Recognition via Weight Balancingcited by 107
  9. 2021The Majority Can Help The Minority: Context-rich Minority Oversampling for Long-tailed Classificationcited by 100
  10. 2021Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognitioncited by 95
  11. 2021Influence-Balanced Loss for Imbalanced Visual Classificationcited by 91
  12. 2022Balanced MSE for Imbalanced Visual Regressioncited by 89
38 more
  1. 2021FASA: Feature Augmentation and Sampling Adaptation for Long-Tailed Instance Segmentationcited by 76
  2. 2022A Survey on Long-Tailed Visual Recognitioncited by 75
  3. 2021Distilling Virtual Examples for Long-tailed Recognitioncited by 72
  4. 2022LPT: Long-tailed Prompt Tuning for Image Classificationcited by 69
  5. 2021ABC: Auxiliary Balanced Classifier for Class-imbalanced Semi-supervised Learningcited by 65
  6. 2022Nested Collaborative Learning for Long-Tailed Visual Recognitioncited by 63
  7. 2022Expanding Small-Scale Datasets with Guided Imaginationcited by 61
  8. 2023Long-Tail Learning with Foundation Model: Heavy Fine-Tuning Hurtscited by 58
  9. 2024Scaling Wearable Foundation Modelscited by 57
  10. 2021Adaptive Class Suppression Loss for Long-Tail Object Detectioncited by 55
  11. 2023No Fear of Classifier Biases: Neural Collapse Inspired Federated Learning with Synthetic and Fixed Classifiercited by 55
  12. 2022RankSim: Ranking Similarity Regularization for Deep Imbalanced Regressioncited by 52
  13. 2021DASO: Distribution-Aware Semantics-Oriented Pseudo-label for Imbalanced Semi-Supervised Learningcited by 51
  14. 2021Exploring Classification Equilibrium in Long-Tailed Object Detectioncited by 51
  15. 2024A Hard-to-Beat Baseline for Training-free CLIP-based Adaptationcited by 51
  16. 2021Adversarial Robustness under Long-Tailed Distributioncited by 50
  17. 2021A Simple Long-Tailed Recognition Baseline via Vision-Language Modelcited by 49
  18. 2022Equalized Focal Loss for Dense Long-Tailed Object Detectioncited by 46
  19. 2023Understanding Imbalanced Semantic Segmentation Through Neural Collapsecited by 46
  20. 2022Generalized Parametric Contrastive Learningcited by 45
  21. 2022Neural Collapse Inspired Attraction-Repulsion-Balanced Loss for Imbalanced Learningcited by 43
  22. 2023Global and Local Mixture Consistency Cumulative Learning for Long-tailed Visual Recognitionscited by 43
  23. 2024LLM as Dataset Analyst: Subpopulation Structure Discovery with Large Language Modelcited by 43
  24. 2024Probabilistic Contrastive Learning for Long-Tailed Visual Recognitioncited by 42
  25. 2022TAM: Topology-Aware Margin Loss for Class-Imbalanced Node Classificationcited by 41
  26. 2021On Model Calibration for Long-Tailed Object Detection and Instance Segmentationcited by 39
  27. 2021Towards Calibrated Model for Long-Tailed Visual Recognition from Prior Perspectivecited by 39
  28. 2022Test-Time Robust Personalization for Federated Learningcited by 39
  29. 2023Eliminating Domain Bias for Federated Learning in Representation Spacecited by 38
  30. 2024Gradient Reweighting: Towards Imbalanced Class-Incremental Learningcited by 37
  31. 2023FedCP: Separating Feature Information for Personalized Federated Learning via Conditional Policycited by 36
  32. 2023How Re-sampling Helps for Long-Tail Learning?cited by 36
  33. 2021Robust Long-Tailed Learning under Label Noisecited by 35
  34. 2023Exploring Vision-Language Models for Imbalanced Learningcited by 35
  35. 2023MosaicFusion: Diffusion Models as Data Augmenters for Large Vocabulary Instance Segmentationcited by 35
  36. 2022Invariant Feature Learning for Generalized Long-Tailed Classificationcited by 34
  37. 2022Learning Imbalanced Data with Vision Transformerscited by 34
  38. 2023Fill-Up: Balancing Long-Tailed Data with Generative Modelscited by 34

50 most cited of 397 All 397 (JSON)