MULTI-LAYERED SPEECH RECOGNITION APPARATUS AND METHOD

• US 20130124197A1
• Filed: 01/02/2013
• Published: 05/16/2013
• Est. Priority Date: 10/08/2004
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Cho, Jeongmi, Hwang, Kwangil, Lee, Yongbeom, Kim, Jeongsu, LEE, Jaewon
• Granted Patent: US 8,892,425 B2
• Time in Patent Office: Days
• US Class Current: 704/201
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems

Abstract

A multi-layered speech recognition apparatus and method, the apparatus includes a client checking whether the client recognizes the speech using a characteristic of speech to be recognized and recognizing the speech or transmitting the characteristic of the speech according to a checked result; and first through N-th servers, wherein the first server checks whether the first server recognizes the speech using the characteristic of the speech transmitted from the client, and recognizes the speech or transmits the characteristic according to a checked result, and wherein an n-th (2≦n≦N) server checks whether the n-th server recognizes the speech using the characteristic of the speech transmitted from an (n−1)-th server, and recognizes the speech or transmits the characteristic according to a checked result.

Claims

1. A method of speech recognition, the method comprising;

obtaining the speech from a user;

characterizing the speech in a client; and

optimally distributing a workload of the speech recognition in the client and servers, based on the characterizing of the speech

2. The method of claim 1, wherein the distributing a workload of the speech recognition comprises determining whether the client recognizes the speech and transmitting the speech that the client itself does not recognize.

3. The method of claim 1, wherein the distributing a workload of the speech recognition in the client and server is performed by the client and a first server, the distributing a workload of the speech recognition in servers is performed by a n-th server, and

wherein the first server has a larger capacity of resource than the client and the n-th server has a larger capacity of resource than the (n−

1)-th server.