keyword spotting


Also found in: Wikipedia.

keyword spotting

[′kē‚wərd ‚spät·iŋ]
(engineering acoustics)
An approach to task-oriented speech understanding through detecting a limited number of keywords that would most likely express the intent of a speaker, rather than attempting to recognize every word in an utterance.
References in periodicals archive ?
Sensory's TrulyHandsfree trigger technology has quickly become the most widely adopted keyword spotting technology in the speech industry.
CSpotter is a voice recognition engine that supports low-power always listening with keyword spotting.
Fortemedia has completed production of its iM401D low power, keyword spotting voice trigger and control IC.
With an extremely noise robust and accurate solution that responds quickly and at ultra-low power consumption, the TrulyHandsfree[TM] trigger technology has become the most widely adopted keyword spotting technology in the speech industry.
Joe was sure to point out how the company's offering features recording, UC; real-time speech analytics; quality management; multichannel communications including voice, web chat and social media; routing; and real-time speech analytics with keyword spotting.
Social media presents unique challenges to organizations relying on traditional keyword spotting technologies to identify sentiment and concepts in social networks.
0 can be enhanced with features like customer feedback, keyword spotting, transcription and speech mining to provide a detailed and automated analysis of recorded calls.
The new features now provide the same functionality as the company's Linux product including recording of encrypted calls, keyword spotting (searching for spoken words), multilingual applications and adjustable modes for individual, bulk or selective recording on each IP phone.
com's Developer Sandbox Program to take advantage of Ditech's innovative services such as keyword spotting and whisper messages into new applications.
Traditional approaches to keyword spotting have failed in high noise and frequently false fire, but TrulyHandsfree (TM) shows unparalleled accuracy without false fires even in high noise and speech.
Furthermore, the greater the number of different analytics methodologies that are applied to customer calls (for example, keyword spotting with ad-hoc query capabilities, emotion detection, talk-over analysis, speaker recognition), the easier it is to understand not only what was said, but in what way, by whom and, most important, what the implications are.