Online environmental adaptation of CNN-based acoustic models using spatial diffuseness features

We propose a new concept for adapting CNN-based acoustic models using spatial diffuseness features as auxiliary information about the acoustic environment: the spatial diffuseness features are simultaneously employed as acoustic-model input features and to estimate environmental cues for context ada...

Full description

Saved in:
Bibliographic Details
Main Authors: Huemmer, Christian, Delcroix, Marc, Ogawa, Atsunori, Kinoshita, Keisuke, Nakatani, Tomohiro, Kellermann, Walter
Format: Conference Proceeding
Language:English
Subjects:
Acoustics
Adaptation models
CNN-based acoustic model
Context
Context adaptation
Context modeling
environmental robustness
Feature extraction
Microphones
spatial diffuseness features
Speech
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We propose a new concept for adapting CNN-based acoustic models using spatial diffuseness features as auxiliary information about the acoustic environment: the spatial diffuseness features are simultaneously employed as acoustic-model input features and to estimate environmental cues for context adaptation, where one convolutional layer is factorized into several sub-layers to represent different acoustic conditions. This context-adaptive CNN-based acoustic model facilitates an online environmental adaptation and is experimentally verified for the real-world recordings provided by the CHiME-3 task. The best performing setup reduces the average word error rate scores achieved by the baseline system (without using spatial diffuseness features) from 19.4% to 15.9% and 12.2% to 10.7% considering two experimental setups with and without front-end signal enhancement, respectively.
ISSN:2379-190X
DOI:10.1109/ICASSP.2017.7953083