S
Shuai Wang
Researcher at Shanghai Jiao Tong University
Publications - 54
Citations - 1281
Shuai Wang is an academic researcher from Shanghai Jiao Tong University. The author has contributed to research in topics: Speaker recognition & Softmax function. The author has an hindex of 15, co-authored 49 publications receiving 696 citations. Previous affiliations of Shuai Wang include Brno University of Technology.
Papers
More filters
Posted Content
BUT System Description to VoxCeleb Speaker Recognition Challenge 2019
TL;DR: The submission of Brno University of Technology (BUT) team to the VoxCeleb Speaker Recognition Challenge (VoxSRC) 2019 is described, a fusion of 4 Convolutional Neural Network (CNN) topologies and the best systems for Fixed and Open conditions achieved 1.42% and 1.26% ERR on the challenge evaluation set respectively.
Proceedings ArticleDOI
Angular Softmax for Short-Duration Text-independent Speaker Verification.
Zili Huang,Shuai Wang,Kai Yu +2 more
TL;DR: In this article, the angular softmax (A-softmax) loss is introduced to improve speaker embedding quality in an end-to-end speaker verification system, where deep discriminant analysis is used for channel compensation.
Proceedings ArticleDOI
Bayesian HMM Based x-Vector Clustering for Speaker Diarization.
TL;DR: This paper presents a simplified version of the previously proposed diarization algorithm based on Bayesian Hidden Markov Models, which uses Variational Bayesian inference for very fast and robust clustering of x-vector (neural network based speaker embeddings).
Journal ArticleDOI
Past review, current progress, and challenges ahead on the cocktail party problem
TL;DR: This overview paper focuses on the speech separation problem given its central role in the cocktail party environment, and describes the conventional single-channel techniques such as computational auditory scene analysis (CASA), non-negative matrix factorization (NMF) and generative models, and the newly developed deep learning-based techniques.
Posted Content
Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognition
TL;DR: Three different margin based losses which not only separate classes but also demand a fixed margin between classes are introduced to deep speaker embedding learning and it could be demonstrated that the margin is the key to obtain more discriminative speaker embeddings.