SPEECH RECOGNITION METHOD AND APPARATUS USING DEVICE INFORMATION

• US 20170206903A1
• Filed: 05/06/2015
• Published: 07/20/2017
• Est. Priority Date: 05/23/2014
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: LEE, Jae-won, KIM, Tae-yoon, HAN, Chang-woo
• Granted Patent: US 10,643,620 B2
• Time in Patent Office: Days
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

A speech recognition method includes: storing at least one acoustic model (AM); obtaining, from a device located outside the ASR server, a device ID for identifying the device; obtaining speech data from the device; selecting an AM based on the device ID; performing speech recognition on the speech data by using the selected AM; and outputting a result of the speech recognition.

Claims

1. A server comprising:

an acoustic model (AM) database configured to store at least one AM;

a reception unit configured to receive, from a device located outside the ASR server, a device ID for identifying the device and speech data; and

a speech recognition unit configured to select an AM based on the device ID, perform speech recognition on the speech data by using the selected AM, and output a result of the speech recognition.

2. The server of claim 1, wherein the AM database stores a general AM and at least one device-adapted AM, and

if a device-adapted AM corresponding to the device ID is stored in the AM database, the speech recognition unit performs speech recognition on the speech data by using the device-adapted AM, and if the device-adapted AM corresponding to the device ID is not stored in the AM database, the speech recognition unit performs speech recognition on the speech data by using the general AM.

3. The server of claim 1, further comprising:

a usage log database configured to store a usage data log that includes the result of the speech recognition and the speech data corresponding to the result of the speech recognition;

a log monitor configured to select a device ID that needs device adaptation, by monitoring the usage log database; and

an AM adaptation unit configured to generate a device-adapted AM corresponding to the selected device ID, by using speech data of the usage data log corresponding to the selected device ID.

4. A speech recognition method comprising:

storing at least one acoustic model (AM);

obtaining, from a device located outside the ASR server, a device ID for identifying the device;

obtaining speech data from the device;

selecting an AM based on the device ID;

performing speech recognition on the speech data by using the selected AM; and

outputting a result of the speech recognition.

5. The speech recognition method of claim 4, wherein the storing of the at least one AM comprises storing a general AM and at least one device-adapted AM, and

the selecting of the AM comprises:

determining whether a device-adapted AM corresponding to the device ID is present; and

selecting the device-adapted AM if the device-adapted AM corresponding to the device ID is present, and selecting the general AM if the device-adapted AM corresponding to the device ID is not present.

6. The speech recognition method of claim 4, further comprising:

storing a usage data log that includes the result of the speech recognition and the speech data corresponding to the result of the speech recognition in a usage log database;

selecting a device ID that needs device adaptation, by monitoring the usage log database; and

generating a device-adapted AM corresponding to the selected device ID, by using speech data of the usage data log corresponding to the selected device ID.

25. A non-transitory computer-readable recording storage medium having stored thereon a computer program, which when executed by a computer, performs the method of claim 4.

7. A device comprising:

a memory configured to store a device ID for identifying a device;

an input unit configured to receive an input of a speech for speech recognition;

a processor configured to generate speech data by processing the speech;

a transmission unit configured to transmit the device ID and the speech data to an automatic speech recognition (ASR) server; and

a reception unit configured to receive a result of the speech recognition performed on the speech data from the ASR server.

8. The device of claim 7, wherein the processor extracts data, which is needed for the speech recognition, as speech data.

9. The device of claim 7, wherein the device has a plurality of device identifications (IDs).

10. The device of claim 9, wherein a setting of the device varies depending on the device IDs.

11. The device of claim 7, wherein the transmission unit transmits location information of the device.

12-24. -24. (canceled)

26. (canceled)