Scholarlink

<p>In this research paper, an innovative Kannada Sign Language interpretation algorithm was designed. This system was intended to assist individuals with hearing impairments in Karnataka in expressing themselves more effectively. Advanced machine learning and gesture detection techniques were employed; however, to enhance throughput and reduce computational overhead, a novel method called ‘spectral correlative coding’ was applied. A comprehensive Kannada Sign Language dataset was created, incorporating 17 distinct sign categories to strengthen recognition capabilities. The performance of the proposed system was evaluated through comparisons with existing methods such as Hidden Markov Models (HMM), Convolutional Neural Networks (CNN), and Support Vector Machines (SVM). It was observed that the proposed approach achieved superior accuracy, particularly in handling grammatical complexities and gesture variations in Kannada Sign Language. This work not only demonstrates significant improvement in regional sign recognition systems, but also enables the full inclusion of Kannada-speaking deaf individuals in education, communication, and employment, thereby promoting societal inclusivity.</p>

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

Enhancing Communication for the Deaf: A Machine Learning Approach to Real-Time Kannada Sign Language Translation

  • 期刊论文

  • 发表日期：

    2025年10月10日

查看全文
  • Ramesh M. Kagalkar,
  • K. Krishna Prasad,
  • Shantappa G. Gollagi

摘要

In this research paper, an innovative Kannada Sign Language interpretation algorithm was designed. This system was intended to assist individuals with hearing impairments in Karnataka in expressing themselves more effectively. Advanced machine learning and gesture detection techniques were employed; however, to enhance throughput and reduce computational overhead, a novel method called ‘spectral correlative coding’ was applied. A comprehensive Kannada Sign Language dataset was created, incorporating 17 distinct sign categories to strengthen recognition capabilities. The performance of the proposed system was evaluated through comparisons with existing methods such as Hidden Markov Models (HMM), Convolutional Neural Networks (CNN), and Support Vector Machines (SVM). It was observed that the proposed approach achieved superior accuracy, particularly in handling grammatical complexities and gesture variations in Kannada Sign Language. This work not only demonstrates significant improvement in regional sign recognition systems, but also enables the full inclusion of Kannada-speaking deaf individuals in education, communication, and employment, thereby promoting societal inclusivity.