An Experimental Study on Noise File Detection using Multiple Frame Classifiers

Punnoose A K

This paper presents an approach to detect noisy wave files, using multiple frame classifiers. The differences in frame classification between two classifiers is codified in terms of probability distributions. A probabilistic framework is presented to detect noisy files using these distributions. The proposed approach is benchmarked against relevant dataset

Select Volume / Issues:

• Volume 9,Issue 3, May 2018

Year:

2018

Type of Publication:

Article

Keywords:

Noise Robustness; Neural Networks; Interactive Voice Response Systems; Phone Duration

Journal:

IJECCE

Volume:

9

Number:

3

Pages:

113-116

Month:

May

ISSN:

2249-071X

Comments:

Paper uploading is in process