Interactive Learning of a Multiple-Attribute Hash Table Classifier for Fast Object Recognition

We consider a parallel, rule-based approach for learning and recognition of pattern and objects in scenes. Classification rules for pattern fragments are learned with objects presented in isolation and are based on unary features of pattern parts and binary features of part relations. These rules are then applied to scenes composed of multiple objects. We present an approach that solves, at the same time, evidence combination and consistency analysis of multiple rule instantiations. Finally, we introduce an extension of our approach to the learning of dynamic patterns.

/pdf/parallel-techniques-for-rule-based-scene-interpretation-2elni77uzo.pdf

Parallel Techniques for Rule-Based Scene Interpretation

MULTI-HASH: learning object attributes and hash tables for fast 3-D object recognition

We have discussed the various schemes used today for representing color, brought out the relationships between them, and, finally, discussed the edge and region based segmentation of color images. We have compared some commonly used segmentation techniques using different representations of color on real data. The report should prove useful as a first reading to students of computer vision interesting in the processing of color images.

/pdf/representation-of-color-and-segmentation-of-color-images-4b9rqoenlx.pdf

Lynne Grewe

Papers

MULTI-HASH: learning object attributes and hash tables for fast 3-D object recognition

Representation of color and segmentation of color images