What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to study the visual similarity of brocades of ethnic minorities in Southwest China through deep - learning methods. Specifically, the paper attempts to solve the following key problems: 1. **Limitations of traditional analysis methods**: - Traditional research on ethnic minority patterns mainly relies on manual analysis. This method is not only inefficient but also easily influenced by subjective biases, and it is difficult to be applied to large - scale and complex data analysis scenarios. 2. **Improving feature extraction and recognition accuracy**: - Existing deep - learning models have certain limitations when dealing with complex ethnic minority patterns, especially in terms of feature extraction. Therefore, the paper proposes an improved convolutional neural network (SResNet - 18) to improve the feature extraction ability and recognition accuracy. 3. **Quantifying and visualizing the similarity between different ethnic patterns**: - The paper hopes to quantify the similarity between different ethnic patterns by introducing three measurement methods (cosine similarity, Euclidean distance, and Manhattan distance) and visualize the results on an ethnic - themed map, so as to more intuitively show the relationships between these patterns and their regional distributions. 4. **Exploring the relationship between cultural patterns and geographical distributions**: - By analyzing the similarity and distribution of different ethnic patterns, the paper reveals the association between cultural patterns and their geographical distributions, providing new perspectives and methods for future research. ### Main contributions - **Proposing a new network architecture (SResNet - 18)**: By embedding the SE module in ResNet - 18, the feature extraction ability and recognition accuracy are significantly improved. - **Achieving high - precision classification performance**: SResNet - 18 reaches an accuracy rate of 98.7% on the test set, which is better than other commonly used models (such as ResNet - 18, VGGNet - 16, and AlexNet). - **Providing multi - dimensional similarity measurement methods**: Using cosine similarity, Euclidean distance, and Manhattan distance to evaluate feature vectors and intuitively showing the results through an ethnic - themed map. - **Revealing the relationship between cultural patterns and geographical distributions**: Through visual analysis, the similarities and differences between different ethnic patterns are shown, as well as their connections with geographical distributions. ### Summary By introducing deep - learning technology, especially the improved convolutional neural network (SResNet - 18), this paper overcomes the limitations of traditional methods in the research of ethnic minority patterns, improves the accuracy of feature extraction and recognition, and reveals the relationship between cultural patterns and geographical distributions through multi - dimensional similarity measurement methods and visualization means. This provides new tools and methods for future research.

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

Self-Adaptive Computational Aesthetic Evaluation of Chinese Ink Paintings Based on Deep Learning

Inkthetics: A Comprehensive Computational Model for Aesthetic Evaluation of Chinese Ink Paintings.

Identification of Miao Embroidery in Southeast Guizhou Province of China Based on Convolution Neural Network

ColorNetVis: an Interactive Color Network Analysis System for Exploring the Color Composition of Traditional Chinese Painting

Segmentation and Synthesis of Embroidery Art Images Based on Deep Learning Convolutional Neural Networks

Assessing similarities between spatial point patterns with a Siamese neural network discriminant model

Research on embroidery image recognition based on deep learning

Do They All Look the Same? Deciphering Chinese, Japanese and Koreans by Fine-Grained Deep Learning

Convolutional neural network application for supply–demand matching in Zhuang ethnic clothing image classification

Visualizing Deep Similarity Networks

Research on Emotion Analysis of Chinese Literati Painting Images Based on Deep Learning

Research on the innovative application of Shen Embroidery cultural heritage based on convolutional neural network

Application of the artificial intelligence system based on graphics and vision in ethnic tourism of subtropical grasslands

Interpreting Regional Characteristics of Tibetan-Qiang Houses in Northwestern Sichuan by Deep Learning and Image Landscape

Image colour application rules of Shanghai style Chinese paintings based on machine learning algorithm

Zhuang Characteristic Culture Detection Based on Improved YOLOV3-SPP Network

The Application of ResNet-34 Model Integrating Transfer Learning in the Recognition and Classification of Overseas Chinese Frescoes

DeepSSN: A deep convolutional neural network to assess spatial scene similarity

Research on Big Data Analysis Technology of Chinese Traditional Culture Yue embroidery Color Network

Intelligent extraction method of ethnic minority decorative patterns in Northeast China based on GABRS algorithm