Gradient weighting for speaker verification in extremely low Signal-to-Noise Ratio.
Yi MaKong Aik LeeVille HautamäkiMeng GeHaizhou LiPublished in: CoRR (2024)
Keyphrases
- low signal to noise ratio
- speaker verification
- random noise
- noisy environments
- prosodic features
- speaker recognition
- audio visual
- emotion recognition
- multilayer perceptron
- edge detection
- language identification
- neural network
- face verification
- using artificial neural networks
- similarity measure
- image processing
- computer vision