mlsanityData to 14 Oct 2026

Large-Margin Softmax Loss for Convolutional Neural Networks

arXiv 1612.02295 ↗First posted 7 Dec 2016stat.ML cs.LGcited by 296

Not on the map. Why?

Abstract (arXiv)

Cross-entropy loss together with softmax is arguably one of the most common used supervision components in convolutional neural networks (CNNs). Despite its simplicity, popularity and excellent performance, the component does not explicitly encourage discriminative learning of features. In this paper, we propose a generalized large-margin softmax (L-Softmax) loss which explicitly encourages intra-class compactness and inter-class separability between learned features. Moreover, L-Softmax not only can adjust the desired margin but also can avoid overfitting. We also show that the L-Softmax loss can be optimized by typical stochastic gradient descent. Extensive experiments on four benchmark datasets demonstrate that the deeply-learned features with L-softmax loss become more discriminative, hence significantly boosting the performance on a variety of visual classification and verification tasks.

Key facts. Cited by 296 papers in the map (data through 14 Oct 2026). First posted on arXiv 7 Dec 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 90; 2022: 71; 2023: 59; 2024: 39; 2025: 24; 2026: 13.

296 papers.

Sort:

By area:

  1. 2021Exploring Cross-Image Pixel Contrast for Semantic Segmentationcited by 242
  2. 2021WebFace260M: A Benchmark Unveiling the Power of Million-Scale Deep Face Recognitioncited by 223
  3. 2022Unified Contrastive Learning in Image-Text-Label Spacecited by 194
  4. 2021Open-World Semi-Supervised Learningcited by 165
  5. 2021ElasticFace: Elastic Margin Loss for Deep Face Recognitioncited by 144
  6. 2022Rethinking Semantic Segmentation: A Prototype Viewcited by 140
  7. 2021Frequency-aware Discriminative Feature Learning Supervised by Single-Center Loss for Face Forgery Detectioncited by 134
  8. 2021Contrastive Learning based Hybrid Networks for Long-Tailed Image Classificationcited by 130
  9. 2022Forward Compatible Few-Shot Class-Incremental Learningcited by 127
  10. 2021Deformable ProtoPNet: An Interpretable Image Classifier Using Deformable Prototypescited by 120
  11. 2023Low-Cost High-Power Membership Inference Attackscited by 107
  12. 2021SynFace: Face Recognition with Synthetic Datacited by 106
38 more
  1. 2022Neural Collapse: A Review on Modelling Principles and Generalizationcited by 95
  2. 2022GMMSeg: Gaussian Mixture based Generative Semantic Segmentation Modelscited by 87
  3. 2022Debiased Learning from Naturally Imbalanced Pseudo-Labelscited by 85
  4. 2022Are All Losses Created Equal: A Neural Collapse Perspectivecited by 78
  5. 2022PatchNet: A Simple Face Anti-Spoofing Framework via Fine-Grained Patch Recognitioncited by 67
  6. 2023Learning with Fantasy: Semantic-Aware Virtual Contrastive Constraint for Few-Shot Class-Incremental Learningcited by 55
  7. 2021Unleashing the Power of Contrastive Self-Supervised Visual Models via Contrast-Regularized Fine-Tuningcited by 50
  8. 2022Neural Collapse with Normalized Features: A Geometric Analysis over the Riemannian Manifoldcited by 50
  9. 2022A Comprehensive Survey on Deep Gait Recognition: Algorithms, Datasets and Challengescited by 45
  10. 2021Orthogonal Projection Losscited by 44
  11. 2021Deep Learning for Micro-expression Recognition: A Surveycited by 43
  12. 2021SphereFace2: Binary Classification is All You Need for Deep Face Recognitioncited by 42
  13. 2022LESS: Label-Efficient Semantic Segmentation for LiDAR Point Cloudscited by 41
  14. 2021Partner-Assisted Learning for Few-Shot Image Classificationcited by 36
  15. 2021Learning with Hyperspherical Uniformitycited by 35
  16. 2022The Principle of Diversity: Training Stronger Vision Transformers Calls for Reducing All Levels of Redundancycited by 35
  17. 2023Online Class Incremental Learning on Stochastic Blurry Task Boundary via Mask and Visual Prompt Tuningcited by 35
  18. 2021Analyzing Overfitting under Class Imbalance in Neural Networks for Image Segmentationcited by 34
  19. 2021Adaptive Logit Adjustment Loss for Long-Tailed Visual Recognitioncited by 33
  20. 2021PML: Progressive Margin Loss for Long-tailed Age Classificationcited by 32
  21. 2021AutoLoss: Automated Loss Function Search in Recommendationscited by 31
  22. 2024Improving Gloss-free Sign Language Translation by Reducing Representation Densitycited by 31
  23. 2021Simple and Robust Loss Design for Multi-Label Learning with Missing Labelscited by 30
  24. 2022Spatio-Temporal Transformer for Dynamic Facial Expression Recognition in the Wildcited by 29
  25. 2021OODformer: Out-Of-Distribution Detection Transformercited by 27
  26. 2021Regular Polytope Networkscited by 24
  27. 2021End2End Occluded Face Recognition by Masking Corrupted Featurescited by 24
  28. 2021Federated Learning for Face Recognition with Gradient Correctioncited by 24
  29. 2021Semi-TCL: Semi-Supervised Track Contrastive Representation Learningcited by 23
  30. 2021SphereFace Revived: Unifying Hyperspherical Face Recognitioncited by 23
  31. 2024FRoundation: Are Foundation Models Ready for Face Recognition?cited by 23
  32. 2021Structure-Aware Face Clustering on a Large-Scale Graph with bf107 Nodescited by 21
  33. 2021MlTr: Multi-label Classification with Transformercited by 21
  34. 2022SFace: Sigmoid-Constrained Hypersphere Loss for Robust Face Recognitioncited by 21
  35. 2022Maximum Class Separation as Inductive Bias in One Matrixcited by 21
  36. 2021Constrained Optimization to Train Neural Networks on Critical and Under-Represented Classescited by 19
  37. 2021A Geometric Perspective towards Neural Calibration via Sensitivity Decompositioncited by 19
  38. 2021Subtype-aware Unsupervised Domain Adaptation for Medical Diagnosiscited by 18

50 most cited of 296 All 296 (JSON)