Dimension-Specific Margins and Element-Wise Gradient Scaling for Enhanced Matryoshka Speaker Embedding
Open Access
- 1 January 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 45473-45487
Abstract
No abstract availableKeywords
Funding Information
- Institute of Information & Communications Technology Planning & Evaluation
- Korean Government [Ministry of Science and ICT (MSIT)] through the Voice Phishing Information Collection and Processing and Development of a Big Data Investigation Support (IITP-2022-0-00653)
This publication has 37 references indexed in Scilit:
- X2-Softmax: Margin adaptive loss function for face recognitionExpert Systems with Applications, 2024
- ECAPA2: A Hybrid Neural Network Architecture and Training Strategy for Robust Speaker Embeddings2023
- HAMFace: Hardness adaptive margin loss for face recognition with various intra-class variationsExpert Systems with Applications, 2023
- Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognition2019
- ArcFace: Additive Angular Margin Loss for Deep Face Recognition2019
- CosFace: Large Margin Cosine Loss for Deep Face Recognition2018
- SphereFace: Deep Hypersphere Embedding for Face Recognition2017
- Deep Residual Learning for Image Recognition2016
- FaceNet: A unified embedding for face recognition and clustering2015
- Learning a Similarity Metric Discriminatively, with Application to Face Verification2005