id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ajema-7671	Iskandarova Sayyora Nurmamatovna; Jurayev Xudoyshukur Utkir ugli	A Survey of Pose-Based Deep Learning Techniques for Student Behavior Recognition in Educational Environments	2025	6	.pdf	application/pdf	2996	169	41	Once keypoints are obtained, deep learning models interpret these skeletal sequences to recognize specific behaviors. This survey organizes the literature into four themes—recurrent sequence models, graph-convolutional frameworks, attention-augmented networks, and transformer-style architectures—compares publicly available classroom datasets, 152 Journal of Engineering, Mechanics and Architecture www. grnjournal.us and outlines current challenges and promising research directions.	cache/ajema-7671.pdf	txt/ajema-7671.txt