Comments on: Youtube as stock house. http://beanstalktalk.com/beanblog/youtube-as-stock-house/ An Ad Giants sponsored adventure. Sun, 05 Jul 2009 02:17:54 +0000 hourly 1 http://wordpress.org/?v=3.0.1 By: Fred http://beanstalktalk.com/beanblog/youtube-as-stock-house/comment-page-1/#comment-14007 Fred Wed, 23 May 2007 02:43:19 +0000 http://beanstalktalk.com/beanblog/youtube-as-stock-house/#comment-14007 Well, I have to say I can just picture that kind of technology being run at CTU by 24's Chloe. I casually followed a few facial recognition businesses after 9/11, and the tech is fascinating. These days there's a huge supply of images. First, there are all those images of ourselves we routinely toss up on the net to any stranger who might want to know what we look like. Then there are the millions of security cameras, cell phone etc. And the stills from videos, especially when good resolution is commonplace, will be a whole new universe. I'd guess such a system works now with stills in the various motor vehicle departments, but the idea of compliling data from the still and videos on the net or in search engine caches is pretty intriguing, all right, not to mention a little scarey. I wonder if, in addition to facial recognition,there's enough of a science built around movement analysis to help with identification. Then there's always vocal recognition. As for a bot to automate automate the dirty work and do the indexing, that's a whole new wrinkle for the accessible net. The silos, however, would probably have to invite them in, I'd guess. I think you're on to something there, and it could be promising or scary depending on the ultimae use. Well, I have to say I can just picture that kind of technology being run at CTU by 24′s Chloe. I casually followed a few facial recognition businesses after 9/11, and the tech is fascinating. These days there’s a huge supply of images. First, there are all those images of ourselves we routinely toss up on the net to any stranger who might want to know what we look like. Then there are the millions of security cameras, cell phone etc. And the stills from videos, especially when good resolution is commonplace, will be a whole new universe.

I’d guess such a system works now with stills in the various motor vehicle departments, but the idea of compliling data from the still and videos on the net or in search engine caches is pretty intriguing, all right, not to mention a little scarey. I wonder if, in addition to facial recognition,there’s enough of a science built around movement analysis to help with identification. Then there’s always vocal recognition.

As for a bot to automate automate the dirty work and do the indexing, that’s a whole new wrinkle for the accessible net. The silos, however, would probably have to invite them in, I’d guess.

I think you’re on to something there, and it could be promising or scary depending on the ultimae use.

]]> By: Capt Perk http://beanstalktalk.com/beanblog/youtube-as-stock-house/comment-page-1/#comment-13917 Capt Perk Tue, 22 May 2007 16:35:22 +0000 http://beanstalktalk.com/beanblog/youtube-as-stock-house/#comment-13917 Well, Fred, talk about inspiration - here's a derivative idea that sprang from so much of your talk about video lately... Some one should spend some R&D time looking into a video search engine - just think of the time that might be saved if we applied the technology of the TSA and Customs Dept. for facial recognition to searching for various image matches among all the digital storehouses? I mean why not send a bot to crawl through the archives and Silo's (oop's product pun for an Ad Giants offering - sorry) out there in digi-land? Certainly we can associate words with images and let this generate an index-able resource for all kinds of applications, not to mention working marketing folks. I'd be interested in comments - off base and bring in the phasers or interesting? Well, Fred, talk about inspiration – here’s a derivative idea that sprang from so much of your talk about video lately…

Some one should spend some R&D time looking into a video search engine – just think of the time that might be saved if we applied the technology of the TSA and Customs Dept. for facial recognition to searching for various image matches among all the digital storehouses? I mean why not send a bot to crawl through the archives and Silo’s (oop’s product pun for an Ad Giants offering – sorry) out there in digi-land? Certainly we can associate words with images and let this generate an index-able resource for all kinds of applications, not to mention working marketing folks.

I’d be interested in comments – off base and bring in the phasers or interesting?

]]>