Publications

Sparse and Shift-invariant Feature Extraction from Non- negative Data

In Proc. of the IEEE International Conference on Audio and Speech Signal Processing (ICASSP)

Publication date: December 18, 2008

Paris Smaragdis, B. Raj, M. Shashanka

In this paper we describe a technique that allows the extraction of multiple local shift-invariant features from analysis of non-negative data of arbitrary dimensionality. Our approach employs a probabilis- tic latent variable model with sparsity constraints. We demonstrate its utility by performing feature extraction in a variety of domains ranging from audio to images and video.

Learn More

Research Area:  Adobe Research iconAI & Machine Learning