ScatterNet Hybrid Deep Learning (SHDL) Network For Object Classification

TL;DR

The paper introduces the SHDL network, a hybrid deep learning architecture combining ScatterNet, unsupervised, and supervised modules, achieving state-of-the-art object classification with improved efficiency and performance on limited data.

Contribution

It presents a novel hybrid deep learning framework with automatically optimized layers, combining ScatterNet features with unsupervised and supervised learning for enhanced object recognition.

Findings

Achieves state-of-the-art classification accuracy on two datasets.

Outperforms unsupervised and semi-supervised methods like GANs.

Demonstrates robustness with smaller training datasets.

Abstract

The paper proposes the ScatterNet Hybrid Deep Learning (SHDL) network that extracts invariant and discriminative image representations for object recognition. SHDL framework is constructed with a multi-layer ScatterNet front-end, an unsupervised learning middle, and a supervised learning back-end module. Each layer of the SHDL network is automatically designed as an explicit optimization problem leading to an optimal deep learning architecture with improved computational performance as compared to the more usual deep network architectures. SHDL network produces the state-of-the-art classification performance against unsupervised and semi-supervised learning (GANs) on two image datasets. Advantages of the SHDL network over supervised methods (NIN, VGG) are also demonstrated with experiments performed on training datasets of reduced size.