While automatic techniques are being developed, it maybe useful to develop manual tagging techniques for ‘random access’ to video. Most sites assign tags to complete videos — that is not much useful because for videos it will be useful to get to specific time intervals where interesting things happen. This could be accomplished by segmenting videos and assigning tags to each segments. Video researchers are doing this using automatic techniques and are making progress. But these techniques are not yet ready. So in the mean time if one could develop manual, or better still semi-automatic, techniques to assign such time-tags to videos, then searching videos will become more effective. There is a great business opportunity in this area.
Fraser Dinnis brought to my attention a company that has taken first step in this direction. See
This is a step in right direction. There is a lot of scope in this area.
http://www.transana.org/
Transana is software for professional researchers who want to analyze digital video or audio data.
Thanks for the posting about veotag.
I found the service easy to use and produced a very professional looking, well tagged product.
I look forward to watching this company grow.
Good luck veotag!!
I was taking a class with Paul Dourish and we talked about the next generation of designed objects. One of the main texts we used in the talk was Shaping Things, which purports that a RFID tagged wine bottle fifteen years from now will not just be a wine bottle.
It will be the wine bottle plus a massive respository of information about the wine bottle. It will be the user experience at the winery website whose URL is printed on the label. It will be the fully recorded and availble history of the bottle from its molding in Italy to its arrival in your kitchen. It will be countless websites devoted to capturing and parsing information about the varietal and the harvest. It will be both a wine bottle and an entry point into a huge collection of metadata about the wine bottle.
Thinking about what Veotag is doing makes me wonder if the same thing is happening to video.