Human Eyes Inspired Recurrent Neural Networks are More Robust Against Adversarial Noises

06/15/2022
by   Minkyu Choi, et al.
0

Compared to human vision, computer vision based on convolutional neural networks (CNN) are more vulnerable to adversarial noises. This difference is likely attributable to how the eyes sample visual input and how the brain processes retinal samples through its dorsal and ventral visual pathways, which are under-explored for computer vision. Inspired by the brain, we design recurrent neural networks, including an input sampler that mimics the human retina, a dorsal network that guides where to look next, and a ventral network that represents the retinal samples. Taking these modules together, the models learn to take multiple glances at an image, attend to a salient part at each glance, and accumulate the representation over time to recognize the image. We test such models for their robustness against a varying level of adversarial noises with a special focus on the effect of different input sampling strategies. Our findings suggest that retinal foveation and sampling renders a model more robust against adversarial noises, and the model may correct itself from an attack when it is given a longer time to take more glances at an image. In conclusion, robust visual recognition can benefit from the combined use of three brain-inspired mechanisms: retinal transformation, attention guided eye movement, and recurrent processing, as opposed to feedforward-only CNNs.

READ FULL TEXT

page 3

page 4

page 5

page 6

page 9

page 10

page 11

page 16

research
03/02/2021

Brain-inspired algorithms for processing of visual data

The study of the visual system of the brain has attracted the attention ...
research
09/05/2016

Deep Retinal Image Understanding

This paper presents Deep Retinal Image Understanding (DRIU), a unified f...
research
03/16/2021

Bio-inspired Robustness: A Review

Deep convolutional neural networks (DCNNs) have revolutionized computer ...
research
03/18/2022

Towards Robust 2D Convolution for Reliable Visual Recognition

2D convolution (Conv2d), which is responsible for extracting features fr...
research
11/17/2020

Vis-CRF, A Classical Receptive Field Model for VISION

Over the last decade, a variety of new neurophysiological experiments ha...
research
10/27/2022

BI AVAN: Brain inspired Adversarial Visual Attention Network

Visual attention is a fundamental mechanism in the human brain, and it i...
research
05/01/2022

DDDM: a Brain-Inspired Framework for Robust Classification

Despite their outstanding performance in a broad spectrum of real-world ...

Please sign up or login with your details

Forgot password? Click here to reset