Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach
In this study, we designed a biomimetic artificial visual system (AVS) inspired by biological visual system that can process RGB images. Our approach begins by mimicking the photoreceptor cone cells to simulate the initial input processing followed by a learnable dendritic neuron model to replicate...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Biomimetics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-7673/10/1/11 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588956760276992 |
|---|---|
| author | Tianqi Chen Yuki Todo Zhiyu Qiu Yuxiao Hua Delai Qiu Xugang Wang Zheng Tang |
| author_facet | Tianqi Chen Yuki Todo Zhiyu Qiu Yuxiao Hua Delai Qiu Xugang Wang Zheng Tang |
| author_sort | Tianqi Chen |
| collection | DOAJ |
| description | In this study, we designed a biomimetic artificial visual system (AVS) inspired by biological visual system that can process RGB images. Our approach begins by mimicking the photoreceptor cone cells to simulate the initial input processing followed by a learnable dendritic neuron model to replicate ganglion cells that integrate outputs from bipolar and horizontal cell simulations. To handle multi-channel integration, we utilize a nonlearnable dendritic neuron model to simulate the lateral geniculate nucleus (LGN), which consolidates outputs across color channels, an essential function in biological multi-channel processing. Cross-validation experiments show that AVS demonstrates strong generalization across varied object–background configurations, achieving accuracy where traditional models like EfN-B0, ResNet50, and ConvNeXt typically fall short. Additionally, our results across different training-to-testing data ratios reveal that AVS maintains over 96% test accuracy even with limited training data, underscoring its robustness in low-data scenarios. This demonstrates the practical advantage of the AVS model in applications where large-scale annotated datasets are unavailable or expensive to curate. This AVS model not only advances biologically inspired multi-channel processing but also provides a practical framework for efficient, integrated visual processing in computational models. |
| format | Article |
| id | doaj-art-2475459998cd4e8a91189113c8f96e17 |
| institution | Kabale University |
| issn | 2313-7673 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomimetics |
| spelling | doaj-art-2475459998cd4e8a91189113c8f96e172025-01-24T13:24:35ZengMDPI AGBiomimetics2313-76732024-12-011011110.3390/biomimetics10010011Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic ApproachTianqi Chen0Yuki Todo1Zhiyu Qiu2Yuxiao Hua3Delai Qiu4Xugang Wang5Zheng Tang6Division of Electrical Engineering and Computer Science, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Ishikawa, JapanFaculty of Electrical, Information and Communication Engineering, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Ishikawa, JapanDivision of Electrical Engineering and Computer Science, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Ishikawa, JapanDivision of Electrical Engineering and Computer Science, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Ishikawa, JapanBrain Science Institute, Jilin Street, Jilin 132013, ChinaBeijing Euler Cognitive Intelligence Technology Co., Ltd., Beijing 100097, ChinaInstitute of AI for Industries, Chinese Academy of Sciences Nanjing, 168, Tianquan Road, Nanjing 211135, ChinaIn this study, we designed a biomimetic artificial visual system (AVS) inspired by biological visual system that can process RGB images. Our approach begins by mimicking the photoreceptor cone cells to simulate the initial input processing followed by a learnable dendritic neuron model to replicate ganglion cells that integrate outputs from bipolar and horizontal cell simulations. To handle multi-channel integration, we utilize a nonlearnable dendritic neuron model to simulate the lateral geniculate nucleus (LGN), which consolidates outputs across color channels, an essential function in biological multi-channel processing. Cross-validation experiments show that AVS demonstrates strong generalization across varied object–background configurations, achieving accuracy where traditional models like EfN-B0, ResNet50, and ConvNeXt typically fall short. Additionally, our results across different training-to-testing data ratios reveal that AVS maintains over 96% test accuracy even with limited training data, underscoring its robustness in low-data scenarios. This demonstrates the practical advantage of the AVS model in applications where large-scale annotated datasets are unavailable or expensive to curate. This AVS model not only advances biologically inspired multi-channel processing but also provides a practical framework for efficient, integrated visual processing in computational models.https://www.mdpi.com/2313-7673/10/1/11artificial visual systemneural networkdendritic neuronmotion direction detectiondeep learning |
| spellingShingle | Tianqi Chen Yuki Todo Zhiyu Qiu Yuxiao Hua Delai Qiu Xugang Wang Zheng Tang Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach Biomimetics artificial visual system neural network dendritic neuron motion direction detection deep learning |
| title | Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach |
| title_full | Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach |
| title_fullStr | Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach |
| title_full_unstemmed | Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach |
| title_short | Learning Dendritic-Neuron-Based Motion Detection for RGB Images: A Biomimetic Approach |
| title_sort | learning dendritic neuron based motion detection for rgb images a biomimetic approach |
| topic | artificial visual system neural network dendritic neuron motion direction detection deep learning |
| url | https://www.mdpi.com/2313-7673/10/1/11 |
| work_keys_str_mv | AT tianqichen learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT yukitodo learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT zhiyuqiu learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT yuxiaohua learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT delaiqiu learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT xugangwang learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach AT zhengtang learningdendriticneuronbasedmotiondetectionforrgbimagesabiomimeticapproach |