A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway
Some fundamental visual features have been found to be fully extracted before reaching the cerebral cortex. We focus on direction-selective ganglion cells (DSGCs), which exist at the terminal end of the retinal pathway, at the forefront of the visual system. By utilizing a layered pathway composed o...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-08-01
|
Series: | Mathematics |
Subjects: | |
Online Access: | https://www.mdpi.com/2227-7390/11/17/3732 |
_version_ | 1797582196697989120 |
---|---|
author | Sichen Tao Xiliang Zhang Yuxiao Hua Zheng Tang Yuki Todo |
author_facet | Sichen Tao Xiliang Zhang Yuxiao Hua Zheng Tang Yuki Todo |
author_sort | Sichen Tao |
collection | DOAJ |
description | Some fundamental visual features have been found to be fully extracted before reaching the cerebral cortex. We focus on direction-selective ganglion cells (DSGCs), which exist at the terminal end of the retinal pathway, at the forefront of the visual system. By utilizing a layered pathway composed of various relevant cells in the early stage of the retina, DSGCs can extract multiple motion directions occurring in the visual field. However, despite a considerable amount of comprehensive research (from cells to structures), a definitive conclusion explaining the specific details of the underlying mechanisms has not been reached. In this paper, leveraging some important conclusions from neuroscience research, we propose a complete quantified model for the retinal motion direction selection pathway and elucidate the global motion direction information acquisition mechanism from DSGCs to the cortex using a simple spiking neural mechanism. This mechanism is referred to as the artificial visual system (AVS). We conduct extensive testing, including one million sets of two-dimensional eight-directional binary object motion instances with 10 different object sizes and random object shapes. We also evaluate AVS’s noise resistance and generalization performance by introducing random static and dynamic noises. Furthermore, to thoroughly validate AVS’s efficiency, we compare its performance with two state-of-the-art deep learning algorithms (LeNet-5 and EfficientNetB0) in all tests. The experimental results demonstrate that due to its highly biomimetic design and characteristics, AVS exhibits outstanding performance in motion direction detection. Additionally, AVS possesses biomimetic computing advantages in terms of hardware implementation, learning difficulty, and parameter quantity. |
first_indexed | 2024-03-10T23:17:29Z |
format | Article |
id | doaj.art-81dd83756b16406bae795803317b7894 |
institution | Directory Open Access Journal |
issn | 2227-7390 |
language | English |
last_indexed | 2024-03-10T23:17:29Z |
publishDate | 2023-08-01 |
publisher | MDPI AG |
record_format | Article |
series | Mathematics |
spelling | doaj.art-81dd83756b16406bae795803317b78942023-11-19T08:31:23ZengMDPI AGMathematics2227-73902023-08-011117373210.3390/math11173732A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective PathwaySichen Tao0Xiliang Zhang1Yuxiao Hua2Zheng Tang3Yuki Todo4Faculty of Engineering, University of Toyama, Toyama-shi 930-8555, JapanFaculty of Engineering, University of Toyama, Toyama-shi 930-8555, JapanFaculty of Electrical and Computer Engineering, Kanazawa University Kakuma-Machi, Kanazawa 920-1192, JapanFaculty of Engineering, University of Toyama, Toyama-shi 930-8555, JapanFaculty of Electrical and Computer Engineering, Kanazawa University Kakuma-Machi, Kanazawa 920-1192, JapanSome fundamental visual features have been found to be fully extracted before reaching the cerebral cortex. We focus on direction-selective ganglion cells (DSGCs), which exist at the terminal end of the retinal pathway, at the forefront of the visual system. By utilizing a layered pathway composed of various relevant cells in the early stage of the retina, DSGCs can extract multiple motion directions occurring in the visual field. However, despite a considerable amount of comprehensive research (from cells to structures), a definitive conclusion explaining the specific details of the underlying mechanisms has not been reached. In this paper, leveraging some important conclusions from neuroscience research, we propose a complete quantified model for the retinal motion direction selection pathway and elucidate the global motion direction information acquisition mechanism from DSGCs to the cortex using a simple spiking neural mechanism. This mechanism is referred to as the artificial visual system (AVS). We conduct extensive testing, including one million sets of two-dimensional eight-directional binary object motion instances with 10 different object sizes and random object shapes. We also evaluate AVS’s noise resistance and generalization performance by introducing random static and dynamic noises. Furthermore, to thoroughly validate AVS’s efficiency, we compare its performance with two state-of-the-art deep learning algorithms (LeNet-5 and EfficientNetB0) in all tests. The experimental results demonstrate that due to its highly biomimetic design and characteristics, AVS exhibits outstanding performance in motion direction detection. Additionally, AVS possesses biomimetic computing advantages in terms of hardware implementation, learning difficulty, and parameter quantity.https://www.mdpi.com/2227-7390/11/17/3732neural networkspattern recognitionmotion direction detectionretinal direction-selective ganglion cells |
spellingShingle | Sichen Tao Xiliang Zhang Yuxiao Hua Zheng Tang Yuki Todo A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway Mathematics neural networks pattern recognition motion direction detection retinal direction-selective ganglion cells |
title | A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway |
title_full | A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway |
title_fullStr | A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway |
title_full_unstemmed | A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway |
title_short | A Novel Artificial Visual System for Motion Direction Detection with Completely Modeled Retinal Direction-Selective Pathway |
title_sort | novel artificial visual system for motion direction detection with completely modeled retinal direction selective pathway |
topic | neural networks pattern recognition motion direction detection retinal direction-selective ganglion cells |
url | https://www.mdpi.com/2227-7390/11/17/3732 |
work_keys_str_mv | AT sichentao anovelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT xiliangzhang anovelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT yuxiaohua anovelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT zhengtang anovelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT yukitodo anovelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT sichentao novelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT xiliangzhang novelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT yuxiaohua novelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT zhengtang novelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway AT yukitodo novelartificialvisualsystemformotiondirectiondetectionwithcompletelymodeledretinaldirectionselectivepathway |