High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

A Survey on an Efficient Visual Information based Speech Recognition

Author(s):

Sricharan S , Sapthagiri College of Engineering; Veena K R, Sapthagiri College of Engineering; Sharath Chandra, Sapthagiri College of Engineering; Prithvi Simha, Sapthagiri College of Engineering; Sunil R, Sapthagiri College of Engineering

Keywords:

CNN (Convoluted Neural Networks) LSTM (Long Short-Term Memory), OOV (Out-Of-Vocabulary), ROC (Receiver Operating Characteristic Curves), ROI (Region of Interest)

Abstract

The survey based on the Efficient Visual Information Based Speech Recognition discloses several aspects about the hidden difficulties in human lipreading. It compares the Face Detection Algorithms, Techniques of Speech Reconstruction, Decoding and Language Model Integration and various Lip-Reading Techniques to provide an insight into building a more efficient lip-reading system and also an efficient method of speech reconstruction methods, by taking into consideration, all the drawbacks of the existing systems.

Other Details

Paper ID: IJSRDV7I30449
Published in: Volume : 7, Issue : 3
Publication Date: 01/06/2019
Page(s): 679-681

Article Preview

Download Article