Document is current

Article History

Received: 3 September 2024

Accepted: 13 February 2025

First Online: 20 February 2025

Declarations

:

: The authors declare no competing interests.

: The research detailed herein adheres strictly to ethical principles in conducting research involving human facial expressions and emotion recognition. To ensure the utmost respect for privacy and confidentiality, the research relies on publicly available datasets for its analysis. Specifically, The datasets used for developing and evaluating the proposed fine-grained human facial key feature extraction and fusion methods include FER-2013, JAFFE, and self-built emotion datasets. Among these, the FER-2013 and JAFFE emotion datasets are widely recognized by the research community and have been used for academic purposes, ensuring that they do not reveal any personal privacy. The data referenced in this study does not contain any real or identifiable information about individuals, organizations, or specific individuals’ emotional states. All facial images and emotional labels within these datasets are de-identified and publicly accessible, eliminating any potential risk of invasion of privacy or unethical use of personal data. The study was performed according to the Declaration of Helsinki. The experimental protocols were approved by the Gansu Provincial Department of Transportation, the Gansu Provincial Department of Science and Technology, and the Academic Ethic Committee of Lanzhou Jiaotong University. All participants involved in the self-built dataset provided informed consent before participating in the experiment. The process of constructing the self-built dataset involves acquiring, storing, managing, interpreting, analyzing, and applying data in an ethical and socially responsible manner. No personally identifiable information was collected. Participants had the right to skip or refuse to answer any questions, and there were no negative consequences for refusing or withdrawing from the experiment. Every effort was made by the researchers to maintain the anonymity and confidentiality of the participants’ data and ensure their security. Furthermore, the research objectives and methods have been designed to advance the field of affective computing and human–computer interaction without any intention to harm or unfairly benefit any individuals or entities. The study aims to enhance the accuracy and robustness of emotion recognition models for broader scientific and technological advancements, adhering to the highest ethical standards in research. In line with ethical practices, this research does not involve the collection of any confidential data or infringe upon privacy rights. The methodology employed, including image preprocessing, feature extraction, and model construction, has been designed to uphold principles of fairness, integrity, and avoidance of conflicts of interest. As this research is purely theoretical and academic in nature, it does not have any immediate real-world implications that could cause harm or unintended consequences. The findings and insights generated from this study contribute to the advancement of knowledge in the field, promoting responsible and ethical practices in the development of emotion recognition technologies.