Unlock instant, AI-driven research and patent intelligence for your innovation.

Geotagged ambient audio for enhanced speech recognition accuracy

A geo-tagging and speech recognition technology, applied in the field of speech recognition, can solve problems such as difficulty in accurately identifying spoken utterances, and achieve the effects of improving the accuracy of speech recognition, increasing computing efficiency, and improving process optimization.

Active Publication Date: 2016-06-08
GOOGLE LLC
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Ambient audio may partially obscure the user's speech making it difficult for an automated speech recognition ("ASR") engine to accurately recognize spoken words

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Geotagged ambient audio for enhanced speech recognition accuracy
  • Geotagged ambient audio for enhanced speech recognition accuracy
  • Geotagged ambient audio for enhanced speech recognition accuracy

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] figure 1 is a diagram of an example system 100 that uses geotagged ambient audio to enhance speech recognition accuracy. figure 1 Also illustrated is the flow of data within the system 100 during state (a) through state (i) and the user interface 158 displayed on the mobile device 104 during state (i).

[0019] More specifically, system 100 includes server 106 and ASR engine 108 in communication with mobile client communication devices including mobile device 102 and mobile device 104 over one or more networks 110 . Server 106 may be a search engine, a dictation engine, a dialogue system, or any other engine or system that uses transcribed speech. The network 110 may include a wireless cellular network, a wireless local area network (WLAN) or a Wi-Fi network, a third generation (3G) or fourth generation (4G) mobile telecommunications network, a private network (such as an intranet), a public network (such as the Internet) or any suitable combination thereof.

[002...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for enhancing speech recognition accuracy. In one aspect, a method includes: receiving geotagged audio signals corresponding to ambient audio recorded by a plurality of mobile devices at a plurality of geographic locations; receiving audio signals corresponding to utterances recorded by a particular mobile device; A specific geographic location associated with a specific mobile device; generating a noise model for the specific geographic location using a subset of the geotagged audio signals, wherein noise is performed on the audio signal corresponding to the utterance using the noise model already generated for the specific geographic location compensate.

Description

[0001] Cross References to Related Applications [0002] This application claims priority to US Application No. 12 / 760,147, filed April 14, 2010, and entitled GEOTAGGEDENVIRONMENTALAUDIOFORENHANCEDSPEECHRECOGNITIONACCURACY, the disclosure of which is incorporated herein by reference. technical field [0003] This specification deals with speech recognition. Background technique [0004] As used in this specification, a "search query" includes one or more query terms that a user submits to a search engine when the user requests the search engine to perform a search query, where a "search term" or "query term" includes one or more A full or partial word, character, or string. The "results" of a search query (or "search results") include, among other things, a Uniform Resource Identifier (URI) that references a resource that the search engine determined was responsive to the search query. The search results may include other things such as titles, preview images, user rating...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G10L21/0208
CPCG10L21/0208G10L15/20
Inventor T·克里斯特詹森M·I·洛伊德
Owner GOOGLE LLC