mlsanityData to 14 Oct 2026

Real Additive Margin Softmax for Speaker Verification

arXiv 2110.09116 ↗First posted 18 Oct 2021cs.SD cs.LG eess.AScited by 3

Not on the map. Why?

Abstract (arXiv)

The additive margin softmax (AM-Softmax) loss has delivered remarkable performance in speaker verification. A supposed behavior of AM-Softmax is that it can shrink within-class variation by putting emphasis on target logits, which in turn improves margin between target and non-target classes. In this paper, we conduct a careful analysis on the behavior of AM-Softmax loss, and show that this loss does not implement real max-margin training. Based on this observation, we present a Real AM-Softmax loss which involves a true margin function in the softmax training. Experiments conducted on VoxCeleb1, SITW and CNCeleb demonstrated that the corrected AM-Softmax loss consistently outperforms the original one. The code has been released at https://gitlab.com/csltstu/sunine.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 18 of its 25 references that are in the map. First posted on arXiv 18 Oct 2021.

Builds on

18 of 25 references.

References · 18

  1. 2018VoxCeleb2: Deep Speaker Recognitioncited by 875
  2. 2017Generalized End-to-End Loss for Speaker Verificationcited by 293
  3. 2018Additive Margin Softmax for Face Verificationcited by 278
  4. 2020In defence of metric learning for speaker recognitioncited by 126
  5. 2020Speaker Recognition Based on Deep Learning: An Overviewcited by 59
  6. 2017Deep Speaker: an End-to-End Neural Speaker Embedding Systemcited by 58
  7. 2019BUT System Description to VoxCeleb Speaker Recognition Challenge 2019cited by 54
  8. 2015End-to-End Text-Dependent Speaker Verificationcited by 49
  9. 2018Exploring the Encoding Layer and Loss Function in End-to-End Speaker and Language Recognition Systemcited by 47
  10. 2019Large Margin Softmax Loss for Speaker Verificationcited by 29
  11. 2019Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognitioncited by 27
  12. 2020CN-Celeb: multi-genre speaker recognitioncited by 27
6 more
  1. 2019Centroid-based deep metric learning for speaker recognitioncited by 20
  2. 2016TristouNet: Triplet Loss for Speaker Turn Embeddingcited by 18
  3. 2018Unified Hypersphere Embedding for Speaker Recognitioncited by 8
  4. 2020Deep Normalization for Speaker Vectorscited by 4
  5. 2015Max-margin Metric Learning for Speaker Recognitioncited by 1
  6. 2018Gaussian-Constrained training for speaker verificationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 0; 2026: 2.

3 papers.

  1. 2026Hyperbolic Additive Margin Softmax with Hierarchical Information for Speaker Verificationcited by 1
  2. 2024Memory-Efficient Training for Deep Speaker Embedding Learning in Speaker Verificationcited by 0
  3. 2026Spoken Language Identification with Pre-trained Models and Margin Losscited by 0