| Description: | Given a large set of photographs, automatically select a smaller subset that contains the best photographs in terms of technical quality and aesthetics, is representative and avoids duplicates or near duplicates. Use existing neural network and classical models to evaluate the image quality and image similarity. Allow the user to influence the selection by choosing suitable parameters. Allow interoperability by using standard image stored in the image files. Allow learning individual user preferences. Compare the performance with existing approaches. Optionally create a web interface to interactively modify the selection.
|
|---|
