PERICASA

This paper presents a novel architecture PERICASA, PERturbed frequent Itemset based classification for Computational Auditory Scene Analysis(CASA). A novel approach for perception of sound waves has been developed. Our aim is to develop a classifier which can correctly identify sound waves from noisy sound mixtures i.e. to solve the classical ‘Cocktail Party Problem’. The architecture is based on Gestalt principles of grouping like Pragnanz, Proximity, Common Fate and Similarity. These grouping cues are incorporated into a new Classification approach which is based on a concept namely Perturbed Frequent Itemsets. The primary idea is more the ease with which we can identify different feature values, easier it is to identify the sound wave.