Audio indexing appeared on Google Labs page
On the page of promising development of Google Labs, a new project has appeared: GAudi (Google Audio Indexing) . This is the technology of recognition and indexing of English speech, which is extracted from multimedia files, including video.
The de facto development of Google began testing two months ago on a small number of videos from the YouTube portal: see the news "A full-text search on video appeared on YouTube . " But it was a kind of “black box”: we could just see how the new feature works, but did not know what really stands behind it. Now a separate interface has been published for searching videos (you can upload any video content from the Internet to this index in the future), as well as a FAQ with information .
From the FAQ, we learned that the speech recognition engine was created from scratch by a special working group of Google employees. Although research in this field has been going on for dozens of years by many companies, but GAudi is a completely independent development of Google.
')
Currently only English is supported and the system, of course, makes a lot of mistakes. For example, in this video the word “Czechoslovakia” is incorrectly recognized as “tech also but there”, and the word “free” is recognized as “forty”, and there are quite a few such errors.
On the project page it is reported that the speech recognition engine will gradually “feed” not only election clips, but also other thematic YouTube channels, and in the long run, probably, it should index video content from other sites as well.
The de facto development of Google began testing two months ago on a small number of videos from the YouTube portal: see the news "A full-text search on video appeared on YouTube . " But it was a kind of “black box”: we could just see how the new feature works, but did not know what really stands behind it. Now a separate interface has been published for searching videos (you can upload any video content from the Internet to this index in the future), as well as a FAQ with information .
From the FAQ, we learned that the speech recognition engine was created from scratch by a special working group of Google employees. Although research in this field has been going on for dozens of years by many companies, but GAudi is a completely independent development of Google.
')
Currently only English is supported and the system, of course, makes a lot of mistakes. For example, in this video the word “Czechoslovakia” is incorrectly recognized as “tech also but there”, and the word “free” is recognized as “forty”, and there are quite a few such errors.
On the project page it is reported that the speech recognition engine will gradually “feed” not only election clips, but also other thematic YouTube channels, and in the long run, probably, it should index video content from other sites as well.
Source: https://habr.com/ru/post/40107/
All Articles