Scholarlink

<p>Face-image is one of the unique biometric measures, can be used for sensitive real time applications in society such as cyber security and forensic investigations. Video streams or image inputs can be used to compare an individual data with known persons’ database for matching in cyber security. Some of the applications of face identification are security, retail and healthcare. The visual cues in the face images are helpful to get more details from the images. Age estimation can be considered to be a non-linear regression problem as the current goal to predict the age in a continuous range of values. The main objective is content browsing to age-restricted. The same work can be used for real time marketing applications, If the user is an adult and using e-commerce applications. This is a requirement for any application that will be used for content security and filtering. Many e-commerce applications are targeting age wise, content wise and geo-location wise. In the proposed system, MORPH dataset that contains images labeled age group is used and hybrid Deep learning (DL) and Large Language Modelling (LLM) algorithms are utilized for recommendation-based applications. Convolution Neural Network (CNN) algorithm is used for image categorization into age group. The combination of CNN, SVM and LLM proved that it is most suitable for expression-based recommendation. Then advanced recommendation algorithm such as collaborative, content based, hybrid and LLM models were used to identify the recommendation accuracy. LLM model gave highest accuracy of 91%.</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

Age estmation and facial expressions features for recommendation systems

  • 期刊论文

  • 发表日期：

    2026年3月13日

查看全文
  • R. G. Ragitha,
  • V. Khanaa,
  • C. Amuthadevi,
  • D. S. Vijayan,
  • B. Karthik

摘要

Face-image is one of the unique biometric measures, can be used for sensitive real time applications in society such as cyber security and forensic investigations. Video streams or image inputs can be used to compare an individual data with known persons’ database for matching in cyber security. Some of the applications of face identification are security, retail and healthcare. The visual cues in the face images are helpful to get more details from the images. Age estimation can be considered to be a non-linear regression problem as the current goal to predict the age in a continuous range of values. The main objective is content browsing to age-restricted. The same work can be used for real time marketing applications, If the user is an adult and using e-commerce applications. This is a requirement for any application that will be used for content security and filtering. Many e-commerce applications are targeting age wise, content wise and geo-location wise. In the proposed system, MORPH dataset that contains images labeled age group is used and hybrid Deep learning (DL) and Large Language Modelling (LLM) algorithms are utilized for recommendation-based applications. Convolution Neural Network (CNN) algorithm is used for image categorization into age group. The combination of CNN, SVM and LLM proved that it is most suitable for expression-based recommendation. Then advanced recommendation algorithm such as collaborative, content based, hybrid and LLM models were used to identify the recommendation accuracy. LLM model gave highest accuracy of 91%.