Can I read Speech Emotion Recognition Using Different Activation Function on EtoBox?

Speech Emotion Recognition Using Different Activation Function by S Kavitha; N Sanjana; K Yogajeeva; S Sathyavathi is a scholarly article available to read on EtoBox.

What is Speech Emotion Recognition Using Different Activation Function about?

Audio is the other form of sound signals having various parameters. The speech recognition is becoming most popular in all most all the fields like alexa, cortana, siri. Speech recognition has been one of the significant aspects of digital processing and has some applications that involve audio mining, customer care applications, security and forensic, and so on. SER is about to have a natural interaction between human beings and machines. The recurrent neural network has enormous success in sequential or time-series tasks. Long short term memory overcame the drawbacks of RNN's vanishing gradient problem. This paper gives an overview of how LSTM works best on SER systems and provides the impact of various activation functions on SER implementation. In the end, the accuracies calculated on the SER system are performed well on our crema-D dataset.

Author

S Kavitha; N Sanjana; K Yogajeeva; S Sathyavathi

Publisher

IEEE

Published

2021

Language

EN