Evaluating the Visual Similarity of Southwest China's Ethnic Minority Brocade Based on Deep Learning

TL;DR

This study uses deep learning, specifically a customized SResNet-18 model, to analyze and visualize the visual similarities of ethnic minority brocade patterns in Southwest China, achieving high accuracy.

Contribution

It introduces a tailored deep learning model for ethnic pattern similarity analysis and visualizes regional pattern relationships on a thematic map.

Findings

SResNet-18 achieved 98.7% accuracy

Feature vectors effectively represented pattern similarities

Visual mapping revealed regional ethnic connections

Abstract

This paper employs deep learning methods to investigate the visual similarity of ethnic minority patterns in Southwest China. A customized SResNet-18 network was developed, achieving an accuracy of 98.7% on the test set, outperforming ResNet-18, VGGNet-16, and AlexNet. The extracted feature vectors from SResNet-18 were evaluated using three metrics: cosine similarity, Euclidean distance, and Manhattan distance. The analysis results were visually represented on an ethnic thematic map, highlighting the connections between ethnic patterns and their regional distributions.