Scholarlink

<p>Facial Expression Recognition (FER) in mentally disabled individuals is a difficult task due to the unique and varied nature of their facial movements and emotional expressions. Facial Expression Recognition (FER) for individuals with cognitive impairments raises ethical concerns, including a lack of informed consent, potential privacy violations, and misinterpretation of emotions due to biased models. These challenges may compromise personal dignity and lead to inappropriate or harmful decisions if not carefully managed. This study introduces a Gray Wolf Optimizer-based Artificial Neural Network (GWObANN) framework to enhance the accuracy of emotion detection in this domain. The proposed method leverages the optimization capabilities of the Gray Wolf Optimizer (GWO) to fine-tune the weights and parameters of the NN, ensuring a more adaptive and robust learning process. By addressing the challenges posed by atypical expressions and improving feature extraction and classification accuracy, the GWObANN achieves significant improvements in recognizing subtle and diverse emotions. Experimental results demonstrate that the optimized network not only enhances recognition performance but also ensures better generalization in real-world scenarios. This work provides an effective solution for FER in mentally disabled individuals, contributing to improved interaction and emotional understanding in assistive systems.</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

Facial Expression Analysis for Understanding Emotional States in Cognitively Impaired Individuals

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • Ganesh Chandra B. Akoliya,
  • Vidya Sarode,
  • Ravindra Navanath Duche,
  • Surendra P. Ramteke

摘要

Facial Expression Recognition (FER) in mentally disabled individuals is a difficult task due to the unique and varied nature of their facial movements and emotional expressions. Facial Expression Recognition (FER) for individuals with cognitive impairments raises ethical concerns, including a lack of informed consent, potential privacy violations, and misinterpretation of emotions due to biased models. These challenges may compromise personal dignity and lead to inappropriate or harmful decisions if not carefully managed. This study introduces a Gray Wolf Optimizer-based Artificial Neural Network (GWObANN) framework to enhance the accuracy of emotion detection in this domain. The proposed method leverages the optimization capabilities of the Gray Wolf Optimizer (GWO) to fine-tune the weights and parameters of the NN, ensuring a more adaptive and robust learning process. By addressing the challenges posed by atypical expressions and improving feature extraction and classification accuracy, the GWObANN achieves significant improvements in recognizing subtle and diverse emotions. Experimental results demonstrate that the optimized network not only enhances recognition performance but also ensures better generalization in real-world scenarios. This work provides an effective solution for FER in mentally disabled individuals, contributing to improved interaction and emotional understanding in assistive systems.