Artificial Intelligence Based Automatic Speech Emotion and Drunkenness Detection Using Convolutional Neural Network Voice Segment Dropout Optimizer

Authors

  • T. Geetha Department of Computer Science and Engineering, Vinayaka Mission’s Kirupananda Variyar Engineering College, Vinayaka Missions Research Foundation (Deemed to be University), Salem, India.
  • M. Annamalai Department of Computer Science and Engineering, Vinayaka Mission’s Kirupananda Variyar Engineering College, Vinayaka Missions Research Foundation (Deemed to be University), Salem, India.
  • R. J. Vigneshwaran Department of Computer Science and Engineering, Vinayaka Mission’s Kirupananda Variyar Engineering College, Vinayaka Missions Research Foundation (Deemed to be University), Salem, India.
  • A. Yuvaraj Department of Computer Science and Engineering, Vinayaka Mission’s Kirupananda Variyar Engineering College, Vinayaka Missions Research Foundation (Deemed to be University), Salem, India.

DOI:

https://doi.org/10.21276/apjhs.2022.9.4S.20

Keywords:

Artificial intelligence, CNN, Drunkenness recognition, Speech emotion, Voiced segment dropout optimizer

Abstract

Speech recognition is far from an easy task, because of the unique nature of the human emotional text and a speech recognition model must strike a balance between being accurate. It should be accurate and sufficient enough to cause a broad group of datasets to separate and prevent errors. The models of the states created using artificial intelligence: Positive, sadly angry, and likely to fall are the easy ways to find the drunkenness or non-drunkenness. By building on the previous methods of emotional detection, drunkenness data set experiences can accurately identify these states. The solution development goal is to evaluate in real time because of the different voice recognition from consumers. It is reported to give the user an idea of what emotion they are experiencing or whether they are stumbling, if an assistant detects a drunk person, it will alert him or her about or making large online data extracted drunk information. If the model proves reliable, it can be used in voice-activated, the artificial intelligence (AI) is characterized by a lack of social or situational awareness of the person with it interacts to active on first stage. The second stage to direct on support Artificial Intelligence and Convolutional Neural Network Voice Segment Dropout Optimizer (CNN-VSDO) algorithm is a system that maintains the issues mentioned above that minimize the risk of overlapping. The convolutional neural network for classification employs filters to features from the spectrogram. This approach is used in personal assistant systems to provide the highest and most suitable of AI-human interaction and more appropriate state-based interactions between AI and humans.

Downloads

Download data is not yet available.

Downloads

Published

2022-06-25

How to Cite

T. Geetha, M. Annamalai, R. J. Vigneshwaran, & A. Yuvaraj. (2022). Artificial Intelligence Based Automatic Speech Emotion and Drunkenness Detection Using Convolutional Neural Network Voice Segment Dropout Optimizer. Asian Pacific Journal of Health Sciences, 9(4), 103–108. https://doi.org/10.21276/apjhs.2022.9.4S.20