Paper Info
Title
An Investigation Into Speaker Informed Dnn Front-End For Lvcsr
Abstract
Deep Neural Network (DNN) has become a standard method in many ASR tasks. Recently there is considerable interest in "informed training" of DNNs, where DNN input is augmented with auxiliary codes, such as i-vectors, speaker codes, speaker separation bottleneck (SSBN) features, etc. This paper compares different speaker informed DNN training methods in LVCSR task. We discuss mathematical equivalence between speaker informed DNN training and "bias adaptation" which uses speaker dependent biases, and give detailed analysis on influential factors such as dimension, discrimination and stability of auxiliary codes. The analysis is supported by experiments on a meeting recognition task using bottleneck feature based system. Results show that i-vector based adaptation is also effective in bottleneck feature based system (not just hybrid systems). However all tested methods show poor generalisation to unseen speakers. We introduce a system based on speaker classification followed by speaker adaptation of biases, which yields equivalent performance to an i-vector based system with 10.4% relative improvement over baseline on seen speakers. The new approach can serve as a fast alternative especially for short utterances.
Year
Venue
Keywords
2015
2015 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING (ICASSP)
speech recognition, deep neural network, speaker adaptation, speaker informed training, bias adaptation
Field
DocType
ISSN
Speech processing,Bottleneck,Pattern recognition,Computer science,Speech recognition,Equivalence (measure theory),Speaker recognition,Artificial intelligence,Speaker diarisation,Artificial neural network,Hidden Markov model,Hybrid system
Conference
1520-6149
Citations 
PageRank 
References 
7
0.47
20
Authors
3
Name
Order
Citations
PageRank
Yulan Liu1484.19
Penny Karanasou2556.40
Thomas Hain310514.91