Title
Voice Activity Detection for Transient Noisy Environment Based on Diffusion Nets
Abstract
We address voice activity detection in acoustic environments of transients and stationary noises, which often occur in real-life scenarios. We exploit unique spatial patterns of speech and non-speech audio frames by independently learning their underlying geometric structure. This process is done through a deep encoder-decoder-based neural network architecture. This structure involves an encoder t...
Year
DOI
Venue
2019
10.1109/JSTSP.2019.2909472
IEEE Journal of Selected Topics in Signal Processing
Keywords
Field
DocType
Voice activity detection,Decoding,Transient analysis,Noise measurement,Neural networks,Real-time systems,Feature extraction
Computer vision,Pattern recognition,Computer science,Voice activity detection,Communications system,Exploit,Robustness (computer science),Concatenation,Encoder,Artificial intelligence,Diffusion map,Artificial neural network
Journal
Volume
Issue
ISSN
13
2
1932-4553
Citations 
PageRank 
References 
4
0.45
0
Authors
3
Name
Order
Citations
PageRank
Amir Ivry1101.69
Baruch Berdugo225225.63
Israel Cohen314414.80