Distributed voice recognition system using acoustic feature vector modification
AssigneeQUALCOMM Incorporated
InventorsChienchung Chang, Naren Malayath, Byron Yoshio Yafuso
Life of this patent
Estimating the term…
Abstract
A voice recognition system applies speaker-dependent modification functions to acoustic feature vectors prior to voice recognition pattern matching against a speaker-independent acoustic model. An adaptation engine matches a set of acoustic feature vectors X with an adaptation model to select a speaker-dependent feature vector modification function f( ), which is then applied to X to form a modified set of acoustic feature vectors f(X). Voice recognition is then performed by correlating the modified acoustic feature vectors f(X) with a speaker-independent acoustic model.
Claims8
A remote station apparatus comprising:
- an adaptation model containing acoustic pattern information;
- an adaptation engine configured to perform pattern matching of acoustic feature vectors against the acoustic pattern information to identify a selected feature vector modification function, and configured to apply the selected feature vector modification function to the acoustic feature vectors to produce a set of modified acoustic feature vectors for processing by a voice recognition engine using a central acoustic model larger than the adaptation model;
- a control processor for evaluating the performance of the selected feature vector modification function and adjusting the selected feature vector modification function based on the evaluating;
- and
- a communications interface for communicating the modified acoustic feature vectors to the voice recognition engine.
A method comprising:
- retrieving, from an adaptation model, acoustic pattern information;
- performing, using an adaptation engine, pattern matching of acoustic feature vectors against the acoustic pattern information to identity a selected feature vector modification function;
- applying, by the adaptation engine, the selected feature vector modification function to the acoustic feature vectors to produce a set of modified acoustic feature vectors for processing by a voice recognition engine using a central acoustic model larger than the adaptation model;
- evaluating the performance of the selected feature vector modification function and adjusting the selected feature vector modification function based on the evaluating;
- and
- communicating the modified acoustic feature vectors to the voice recognition engine.
Dependencies are inferred from the claim text.