Sparse Deep Stacking Network for Image Classification

TL;DR

This paper introduces a sparse deep stacking network that leverages sparse coding and a novel sparse neural network module to improve image classification accuracy efficiently.

Contribution

It proposes a sparse SNNM module with mixed-norm regularization and stacks these modules to form a deep network for enhanced image classification.

Findings

Achieved 98.8% accuracy on 15 scene database.

Outperformed related methods with only a linear classifier.

Validated on four diverse image datasets.

Abstract

Sparse coding can learn good robust representation to noise and model more higher-order representation for image classification. However, the inference algorithm is computationally expensive even though the supervised signals are used to learn compact and discriminative dictionaries in sparse coding techniques. Luckily, a simplified neural network module (SNNM) has been proposed to directly learn the discriminative dictionaries for avoiding the expensive inference. But the SNNM module ignores the sparse representations. Therefore, we propose a sparse SNNM module by adding the mixed-norm regularization (l1/l2 norm). The sparse SNNM modules are further stacked to build a sparse deep stacking network (S-DSN). In the experiments, we evaluate S-DSN with four databases, including Extended YaleB, AR, 15 scene and Caltech101. Experimental results show that our model outperforms related classification methods with only a linear classifier. It is worth noting that we reach 98.8% recognition accuracy on 15 scene.