A. Pentland, CONTENT-BASED INDEXING OF IMAGES AND VIDEO, Philosophical transactions-Royal Society of London. Biological sciences, 352(1358), 1997, pp. 1283-1290
By representing image content using probabilistic models of an object'
s appearance we can obtain semantics-preserving compression of the ima
ge data. Such compact representations of an image's salient features a
llow rapid computer searches of even large image databases. Examples a
re shown for databases of face images, a video of American sign langua
ge (ASL), and a video of facial expressions.