Scholarlink

This study introduces a 3D convolutional neural network (3D-CNN) model for recognizing sign language expressions from video inputs, specifically tailored to assist communication for individuals with speech and hearing disabilities. The model employs spatiotemporal filters to extract both spatial and temporal features, processing detected faces in each frame to accurately classify sign expressions. We demonstrate our work on INCLUDE-50 and IRKSL datasets, obtain an accuracy of 94% on INCLUDE-50 dataset, and achieve 97% of accuracy on IRKSL dataset.

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

Indian Sign Language Recognition at Word Level Using 3D-CNN

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Srushti Nayak,
  • C. Sujatha,
  • Padmashree D. Desai,
  • F. M. Umadevi

摘要

This study introduces a 3D convolutional neural network (3D-CNN) model for recognizing sign language expressions from video inputs, specifically tailored to assist communication for individuals with speech and hearing disabilities. The model employs spatiotemporal filters to extract both spatial and temporal features, processing detected faces in each frame to accurately classify sign expressions. We demonstrate our work on INCLUDE-50 and IRKSL datasets, obtain an accuracy of 94% on INCLUDE-50 dataset, and achieve 97% of accuracy on IRKSL dataset.