Nonnegative OPLS for Supervised Design of Filter Banks: Application to Image and Audio Feature Extraction

Audio or visual data analysis tasks usually have to deal with high-dimensional and non-negative signals. However, most data analysis methods suffer from overfitting and numerical problems when data have more than a few dimensions needing a dimensionality reduction preprocessing. Moreover, interpretability about how and why filters work for audio or visual applications is a desired property, specially when energy or spectral signals are involved. In these cases, due to the nature of these signals, the non-negativity of the filter weights is a desired property to better understand its working. Because of these two necessities, we propose different methods to reduce the dimensionality of data while the non-negativity and interpretability of the solution are assured. In particular, we propose a generalized methodology to design filter banks in a supervised way for applications dealing with non-negative data, and we explore different ways of solving the proposed objective function consisting of a non-negative version of Orthonormalized Partial Least Squares (OPLS) method. We analyze the discriminative power of the features obtained with the proposed methods for two different and widely studied applications: texture and music genre classification. Furthermore, we compare the filter banks achieved by our methods with other state-of-the-art methods specifically designed for feature extraction.