Document is current

Article History

Received: 27 November 2025

Accepted: 16 March 2026

First Online: 28 March 2026

Declarations

:

: The authors declare no competing interests.

: The benchmark datasets used in this research are openly available public datasets created and released by their respective authors for academic use. They do not contain identifiable human data requiring ethical review, and the authors on human subjects using these datasets performed no experiments. For the Custom Six-Class Dataset, the following points are confirmed: All methods involving human participants were carried out in accordance with relevant guidelines and regulations, Written informed consent was obtained from all subjects for participation, Written informed consent was also obtained for publication of non-identifying visual data in an open-access journal, No identifiable or sensitive personal information is included in this dataset. As the custom dataset consists of simple controlled video recordings without clinical intervention or sensitive data, formal institutional ethical committee approval was not required under our institution’s guidelines. However, all standard ethical practices regarding consent, privacy, and responsible data handling were strictly followed. This study is based on a retrospective analysis of non-identifiable video data obtained from publicly available research datasets and a custom-created dataset. As the research does not involve clinical procedures, interventions, or identifiable human subjects, the Islamia College University Peshawar Ethics Committee, Pakistan, waived the requirement for formal ethical approval. The waiver was granted because the study involves only non-sensitive behavioral video recordings intended for academic research and development of machine learning models. All methods and procedures were carried out in accordance with institutional guidelines and relevant international regulations.