On Parameter Adaptation in Softmax-Based Cross-Entropy Loss for Improved Convergence Speed and Accuracy in DNN-Based Speaker Recognition.
Magdalena RybickaKonrad KowalczykPublished in: INTERSPEECH (2020)
Keyphrases
- convergence speed
- cross entropy
- speaker recognition
- particle swarm optimization algorithm
- particle swarm optimization
- convergence rate
- differential evolution
- vector quantization
- gaussian mixture model
- learning rate
- pso algorithm
- log likelihood
- speaker identification
- probabilistic neural network
- speaker verification
- maximum likelihood
- evaluation metrics
- error function
- training process
- speech signal