No, Google Image Search is not CBIR. Google relies on textual metadata for its search. The "content" in CBIR is the image content - raw pixel values and all that can be learned from them (color, texture, intensity, orientation, etc.). Common CBIR queries are performed by example (providing a sample image) or by more exotic means - drawing a sketch, specifying what the color histogram might looks like, tracking eye movements, and many others.
Google Image Labeler is also not CBIR. It is an example of semantic annotation.
MyYahoo has been able to drag and drop personalized content (their news and any RSS feeds of your choice) for months now (some features it has for years). It is actually very well done.
Slashdot Mirror
No, Google Image Search is not CBIR. Google relies on textual metadata for its search. The "content" in CBIR is the image content - raw pixel values and all that can be learned from them (color, texture, intensity, orientation, etc.). Common CBIR queries are performed by example (providing a sample image) or by more exotic means - drawing a sketch, specifying what the color histogram might looks like, tracking eye movements, and many others.
Google Image Labeler is also not CBIR. It is an example of semantic annotation.
MyYahoo has been able to drag and drop personalized content (their news and any RSS feeds of your choice) for months now (some features it has for years). It is actually very well done.