Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Voice-on-the-net (VON), VoIP, Internet (IP) Telephony -- Ignore unavailable to you. Want to Upgrade?

To: Stephen B. Temple who wrote (2345)	1/11/1999 11:23:00 AM
From: Stephen B. Temple	Respond to of 3178

Great article if you missed it> Will video indexing kill off the database? By John Taschek, PC Week Online A couple of weeks ago, I wrote about Oracle's plans to kill Microsoft Windows by moving all file system activity to an Oracle8i database. But in a strange "small fish going for the bigger fish" scenario, there are some vendors out there that think their technology indirectly can kill Oracle and other databases. One of them is Virage, which ironically is a development partner with Oracle, IBM and Informix. Virage has developed a video-indexing algorithm that allows users to search through video based on defined criteria. You can see what Virage is up to at video.altavista.com. You can also see what President Clinton was up to because the Virage demo allows users to search through Clinton's 4-hour testimony using normal English keywords. Obviously, video searching has some interesting applications. Video has been moving toward an all-digital format for some time. It's ridiculous to assume, however, that digital media will remain subject to the same archaic constrictions found in the analog format, where you jog the video frames back and forth until somehow you magically arrive at the place you were searching for. Virage, with its Video Cataloger, simply reads in the video stream and extracts relevant metadata information. Each analog video source has tracks used to separate video, audio, timing sequences and captions. The Video Cataloger "ingests" these tracks and produces an indexed digital file format. This index allows users to search through the video nonlinearly. In the Clinton video, Virage indexed the captioning of the video so that users can search based on regular English words. The applications for this technology are clear: Lawyers can search through videotapes, the military can provide more effective training on complex weaponry and systems, and educational institutions can access video to enhance studying. On the corporate side, organizations can create indexed and searchable videos of focus groups, and broadcasters can provide instant video based on easy searches. Wanna replay a video of Sammy Sosa's first home run? Just search the video database, and it will play. Virage recently announced a partnership with IBM that combines IBM's speech-to-text engine and broadcast-industry- specific vocabulary with the Virage Video Cataloger. This will allow broadcasters to search not just on words, but also on laugh tracks, audience classifications and text. So where does the database replacement come in? Most organizations still do not put videos into database servers, although it's a completely viable option. Companies have been using streaming technologies such as those from RealNetworks instead. In fact, the success of RealNetworks has put a damper on competitive database offerings, such as the Oracle Video Server and the SGI WebForce MediaBase. Does anyone even remember what happened to Microsoft Tiger? Both static and dynamic file formats are moving directly to the Web server with no database middleman. So why not move everything to Web servers? Although they're not ready to handle the security and transaction processing requirements found in even the most commonplace database servers, many would argue that it is easier to add transaction support and security to a Web server than it is to fix the object-relational mismatches in a traditional database. The time frame of the death of the database is your call. But I'm not betting on this one.