Effects of Gabor Filters on Classification Performance of CNNs Trained on a Limited Number of Conditions

2024-07-02

2024 International Technical Conference on Circuits/Systems, Computers, and Communications (ITC-CSCC)

AI Summary

This work addresses the challenges of deploying convolutional neural networks (CNNs) on edge devices for robotic vision tasks, where limited training data, constrained model capacity, and poor generalization often hinder performance. To overcome these limitations, the authors propose integrating Gabor filters as a biologically inspired preprocessing module, emulating the visual nervous system’s ability to learn efficiently from sparse visual experience. Through systematic experiments across multiple CNN architectures on a newly constructed multi-view image dataset, the study demonstrates that this approach significantly improves classification accuracy and cross-condition generalization under data-scarce and acquisition-constrained scenarios, while simultaneously reducing model parameter count. The findings offer a novel pathway toward lightweight yet highly generalizable edge-based vision systems.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Learning on the Edge & Model Compression

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

In this study, we propose a technique to improve the accuracy and reduce the size of convolutional neural networks (CNNs) running on edge devices for real-world robot vision applications. CNNs running on edge devices must have a small architecture, and CNNs for robot vision applications involving onsite object recognition must be able to be trained efficiently to identify specific visual targets from data obtained under a limited variation of conditions. The visual nervous system (VNS) is a good example that meets the above requirements because it learns from few visual experiences. Therefore, we used a Gabor filter, a model of the feature extractor of the VNS, as a preprocessor for CNNs to investigate the accuracy of the CNNs trained with small amounts of data. To evaluate how well CNNs trained on image data acquired under a limited variation of conditions generalize to data acquired under other conditions, we created an image dataset consisting of images acquired from different camera positions, and investigated the accuracy of the CNNs that trained using images acquired at a certain distance. The results were compared after training on multiple CNN architectures with and without Gabor filters as preprocessing. The results showed that preprocessing with Gabor filters improves the generalization performance of CNNs and contributes to reducing the size of CNNs.

Problem

• CNN
• limited data
• generalization
• edge devices
• robot vision

Innovation

• Gabor filters
• convolutional neural networks
• limited data training
• edge computing
• visual nervous system