Simon Fraser University

Published purpose

The synthesis of motion data presents a significant opportunity to address key scientific gaps in the study of human activities and interactions, particularly in the context of developing assistive agents to support older adults aging in place. One of the primary challenges in this domain is the scarcity of comprehensive and diverse datasets that accurately capture the intricacies of daily activities and interactions within home environments. Existing datasets often lack the granularity, realism, and size necessary to train AI models effectively, hindering the development and deployment of assistive technologies tailored to the specific needs of older adults. Furthermore, the synthesis of motion data allows augmenting existing datasets or generating entirely new datasets that encompass a broader range of activities and scenarios, thereby enabling more robust and generalizable AI models. By bridging this data gap through AI-driven synthesis, new opportunities can be unlocked for studying human behavior, understanding social dynamics, and ultimately designing more personalized and effective assistive agents to support older adults in maintaining their independence and quality of life while aging in place