Scholarlink

Human Activity Recognition (HAR) is a critical node in fields such as healthcare, sports analysis, and human–computer interaction. Regular human activities can be well recognized with existing models and state-of-the-art but not much work has been done in the field of HAR for physically challenged people. In this paper, we propose a never-existing dataset for physically disabled people and recognition of their basic activities using CNN and Bi-LSTM. The accuracy achieved is 64.91% with precision, recall, and F1-score of 65.86%, 63.15%, and 63.07%, respectively.

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

Human Activity Recognition for the Physically Disabled Using CNN-Bi-LSTM

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Geetanjali Bhola,
  • Dinesh Vishwakarma

摘要

Human Activity Recognition (HAR) is a critical node in fields such as healthcare, sports analysis, and human–computer interaction. Regular human activities can be well recognized with existing models and state-of-the-art but not much work has been done in the field of HAR for physically challenged people. In this paper, we propose a never-existing dataset for physically disabled people and recognition of their basic activities using CNN and Bi-LSTM. The accuracy achieved is 64.91% with precision, recall, and F1-score of 65.86%, 63.15%, and 63.07%, respectively.