Energy-Based Spherical Sparse Coding

TL;DR

This paper introduces Energy-Based Spherical Sparse Coding (EB-SSC), a novel model that combines convolutional sparse coding with class-specific biases for improved image classification, emphasizing efficiency and interpretability.

Contribution

The paper proposes EB-SSC, a new spherical sparse coding method incorporating learned biases for discriminative classification, and demonstrates its effectiveness in deep layered image classification models.

Findings

Effective for image classification tasks

Improves interpretability of sparse codes

Demonstrates competitive performance

Abstract

In this paper, we explore an efficient variant of convolutional sparse coding with unit norm code vectors where reconstruction quality is evaluated using an inner product (cosine distance). To use these codes for discriminative classification, we describe a model we term Energy-Based Spherical Sparse Coding (EB-SSC) in which the hypothesized class label introduces a learned linear bias into the coding step. We evaluate and visualize performance of stacking this encoder to make a deep layered model for image classification.