A new approach for selective edge enhancement using unsharp masking is presented. This is based on the premise that biological vision and image reproduction share common principle...
Novelty detection involves identifying novel patterns. They are not usually available during training. Even if they are, the data quantity imbalance leads to a low classification ...
Abstract We present an efficient method for importance sampling the product of multiple functions. Our algorithm computes a quick approximation of the product on-the-fly, based on ...
Fingerprint classification reduces the searching time of an automated fingerprint identification system. Since fingerprints have properties of intra-class diversities and inter-cla...
We present a vision-based method that assists human
navigation within unfamiliar environments. Our main contribution
is a novel algorithm that learns the correlation between
use...