Supervised and Unsupervised Feature Extraction from a Cochlear Model for Speech Recognition

reportActive / Technical Report | Accession Number: ADA261523 | Open PDF

Abstract:

We explore the application of a novel classification method that combines supervised and unsupervised training, and compare its performance to various more classical methods. We first construct a detailed high dimensional representation of the speech signal using Lyons cochlear model and then optimally reproduce its dimensionality. The resulting low dimensional projection retains the information needed for robust speech recognition.

Author(s):
Intrator, N. ; Tajchman, G.
Author Organization(s):
BROWN UNIV PROVIDENCE RI INST FOR BRAIN AND NEURAL SYSTEMS
Descriptive Note:
Technical rept.
Pagination:
0011

Security Markings

DOCUMENT & CONTEXTUAL SUMMARY

Distribution:
Approved For Public Release
Distribution Statement:
Approved For Public Release; Distribution Is Unlimited.

RECORD

Collection: TR
Identifying Numbers
Contract/Grant Number(s):
N00014-91-J-1316
Monitor Series:
ONR
 Subject Terms
Joint Capability Areas:
JCA_1.2.1_Training; JCA_1.2.7_Experimentation; JCA_5_Command and Control
Modernization Areas:
AI and Machine Learning
Communities of Interest:
Energy and Power Technologies
Descriptor(s):
*MODELS, *SPEECH RECOGNITION, NEURAL NETS, TRAINING, PERFORMANCE(HUMAN), SPEECH, MOTOR NEURONS, LEARNING, EXTRACTION, SIGNALS, CLASSIFICATION, RECOGNITION
Field(s)/Group(s):
Anatomy and Physiology, Test Facilities, Equipment and Methods, Voice Communications
Keyword(s):
*COCHLEAR MODEL, FEATURE EXTRACTION
Report Date:
1992 Dec 23