MEConformer: Highly representative embedding extractor for speaker verification via incorporating selective convolution into deep speaker encoder.
Qiuyu ZhengZengzhao ChenZhifeng WangHai LiuMengting LinPublished in: Expert Syst. Appl. (2024)
Keyphrases
- speaker verification
- noisy environments
- speaker recognition
- prosodic features
- audio visual
- emotion recognition
- acoustic features
- image processing
- multilayer perceptron
- speaker diarization
- bit rate
- using artificial neural networks
- language identification
- multiscale
- pattern recognition
- facial expressions
- learning algorithm
- video sequences
- feature extraction