, Wright State University
Pages: p. C2
Most of the discussions of some five to ten years ago about market viability of multimedia products still hold true. The key consideration continues to be what portions, if any, of what we consider to be the core research areas within our field have made or are about to make a major market impact.
Take search and retrieval, an area in which we have made significant strides in advancing the state of the art. We've especially made advances in retrieval of objects in a single media type, such as still pictures. Hopefully you agree that the recent techniques have made image retrieval more versatile, with better acuity and accuracy. One would think that this technology would make a dashing complementary toolkit for the happy-clicking camera owner. However, despite the large and ever-expanding market of digital cameras, we haven't seen our research contributions make an impact in any substantial way. Doesn't sound good, does it?
Unfortunately, this is the reality. If you disagree, think about the following: When selecting a camera, an average buyer considers several factors, almost all relating to the hardware (such as the size of the image capture device, how many megapixels it generates, zoom capability, and so on). The image-management software isn't a clear factor or concern. The end result is that we're perfecting the ability to search for the actual content instead of just for surrogate tags, but buyers don't know about this technology and probably don't care.
Maybe it's the nature of our field. Perhaps we'll never see that elusive killer app that we have fantasized about for so long. Maybe our work is bound to make only marginal contributions, instead of opening a new market or significantly expanding an existing one. What do you think?
Questions are plenty and success stories few and far in between. If you have any thoughts on this topic, please write to me at email@example.com. Are there any up-and-coming areas that you feel will make a larger market impact? Are there things that we should be doing differently? Any guesses on what scenarios may be more likely to lead us to success?