The Poetry of Computation
Algorithms have been described as "the poetry of computation." But it's poetry with a practical purpose - as CSL's research in signal and image processing clearly demonstrated.
New adaptive algorithms separated signal components, making it possible to tune-in FM radio stations more effectively - or to separate one bird call from another.
CSL algorithms, based on wavelets and other techniques, set world records for efficiency in the compression of images and video for storage and transmission.
CSL pioneered the use of relevance feedback in image retrieval. This made it possible for users to search for images according to example and similarity of image features, such as color, texture, shape, and structure.
Researchers devised optimal strategies for selecting sampling patterns in the digitization of signals, images, and time-varying scenes. Important applications included communications and magnetic resonance imaging.
Fast Image Reconstruction
Fast algorithms from CSL made it possible to reconstruct computer tomography images at far greater speeds - an increase on the order of 50 to 100. This work was applied to image formation in synthetic aperture radar with similar success.
A Smart Hearing Aid
Researchers developed signal-processing algorithms for a revolutionary new binaural hearing aid that could focus in a desired direction and remove interference, such as noise in a crowded room.