MMASD
收藏资源简介:
MMASD是由特拉华大学创建的多模态数据集,专注于自闭症干预分析。该数据集包含来自32名自闭症儿童的1,315个数据样本,这些样本是从超过100小时的干预录像中提取的。MMASD通过提供光学流、2D和3D骨骼数据以及临床评估结果,旨在帮助研究人员和治疗师理解儿童的认知状态,监控治疗进展,并相应地定制治疗计划。此外,该数据集还启发了解决社会任务的下游应用,如动作质量评估和人际同步估计。
MMASD is a multimodal dataset created by the University of Delaware, focusing on autism intervention analysis. This dataset contains 1,315 data samples from 32 autistic children, which are extracted from over 100 hours of intervention video recordings. By providing optical flow, 2D and 3D skeletal data as well as clinical assessment results, MMASD aims to assist researchers and therapists in understanding children's cognitive states, monitoring treatment progress, and tailoring treatment plans accordingly. Additionally, this dataset has inspired downstream applications for solving social tasks, such as motion quality assessment and interpersonal synchronization estimation.




