Author : R.B Pradeeba 1
Date of Publication :22nd February 2018
Abstract: Emotion is a positive or negative state of a person’s mind which is related to physiological activities. The main objective of the paper is to apply Deep Neural Network (DNN) and k-nearest neighbour (k-NN) in recognition of emotion from speech-especially scary state of mind. The most precise model will be utilized to generate alert signals through the cloud. Health care system and in particular intensive and palliative care is supposed to be the area of application. Utterance level is the main feature for analysis with special emphasis. Raw data collection, conversion of acoustic signals to waveforms, probable emotion classification and recognition using databases existing, creating alert signals through the cloud is the sequence of steps followed. The finding of the paper is a fruitful contribution to the palliative care system.
Reference :
-
- sequential k-nearest neighbor pattern recognition for usable speech classification - Jashmin K Shan, Brett Smolenki , Robert E Yantorno , Ananth N Iyer.
- http://www.vocabulary.com
- http://nbviewer.jupyter.org/github/librosa/librosa/blob/ master/examples/libROSA%20demo.ipynb.
- Emotion Detection Analysis Through Tone Of User: A Durvey- Ankita Dutta Chowdhuri , Sachin Bojewar.
- http://www.neuralnetwork.com