Gaussian RAM: Lightweight Image Classification via Stochastic Retina-Inspired Glimpse and Reinforcement Learning

TL;DR

This paper introduces GDRAM, a lightweight, reinforcement learning-based image classification model inspired by biological vision, which efficiently processes large images by mimicking retinal stochasticity, outperforming traditional CNNs.

Contribution

The paper presents GDRAM, a novel stochastic retina-inspired neural network that improves real-time large-scale image classification performance over conventional CNNs.

Findings

GDRAM outperforms CNNs on large cluttered datasets.

The model effectively mimics biological visual processes.

GDRAM achieves high accuracy with reduced computational load.

Abstract

Previous studies on image classification have mainly focused on the performance of the networks, not on real-time operation or model compression. We propose a Gaussian Deep Recurrent visual Attention Model (GDRAM)- a reinforcement learning based lightweight deep neural network for large scale image classification that outperforms the conventional CNN (Convolutional Neural Network) which uses the entire image as input. Highly inspired by the biological visual recognition process, our model mimics the stochastic location of the retina with Gaussian distribution. We evaluate the model on Large cluttered MNIST, Large CIFAR-10 and Large CIFAR-100 datasets which are resized to 128 in both width and height.