Hearing Software

Humans can understand information from multiple domains for ex: speech, text, and visual. With the advancement of deep learning technology, there has been significant improvement in speech recognition. Recognizing emotion from a speech is an important aspect and with deep learning technology emotion recognition has improved in accuracy and latency. There are still many challenges to improve accuracy. We explored different neural networks to improve the accuracy of emotion recognition. With different architectures, we found multi-model architecture which processes audio spectrograms and corresponding video frames giving emotion prediction.