Scholarlink

This paper presents a hybrid CNN–KAN approach for improved image classification. By integrating CNN’s feature extraction with KAN’s non-linear domain-specific processing, the model demonstrates enhanced accuracy and adaptability across benchmark datasets, highlighting its potential for real-world applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid CNN–KAN Models for Benchmark Image Classification

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gautam Kumar,
  • Gobind Singh,
  • Dhruv Sharma,
  • Jai Kumar Vanjare

摘要

This paper presents a hybrid CNN–KAN approach for improved image classification. By integrating CNN’s feature extraction with KAN’s non-linear domain-specific processing, the model demonstrates enhanced accuracy and adaptability across benchmark datasets, highlighting its potential for real-world applications.