This includes: 1. Integrating the faces in the app 2. Processing them and creating the DB that includes their "normalized" "footprint": -A. Locations of eyes and mouth - so I can match the user's face against) -B. Very low-resolution BW representation (say, 30X30) - so I can compare and find the roughly "most similar face" 3. Code that efficiently finds the 4 most similar faces according to the above 5. Integrating/enhancing image processing lab with the following abilities, implemented efficiently -A. direct access to the picture - get RGB value at (x,y) -B. resize, reshape, BW -C. gaussian blur -D. convertion to Lab color space -E. convolution with mask -F. (option) other available filters 6. Providing the code so that I can easily try several effects, which are combinations of the above for f in *.jpg; do echo $f; convert "$f" -resize 64x64 "$(echo $(echo 'temp/resized_' )""$f)"; done