The Idlab Voxsrc-20 Submission: Large Margin Fine-Tuning and Quality-Aware Score Calibration in DNN Based Speaker Verification.
Jenthe ThienpondtBrecht DesplanquesKris DemuynckPublished in: ICASSP (2021)
Keyphrases
- fine tuning
- speaker verification
- quality aware
- noisy environments
- speaker recognition
- prosodic features
- fine tuned
- emotion recognition
- audio visual
- multilayer perceptron
- learning algorithm
- language identification
- support vector
- neural network
- face verification
- basis functions
- multi modal
- knn
- speaker identification
- artificial neural networks
- image retrieval
- computer vision