GenAI4HS

Human sensing plays a critical role in wide ranging applications such as human activity recognition, health monitoring and behavior analysis, often driven by the development of specialized models. Meanwhile, broadly usable, generic models, particularly involving Generative AI and foundational models, have brought transformative advancements in other domains. Yet, similar breakthroughs have not yet been widely observed in wearable and ubiquitous computing.

This workshop aims to bring researchers and practitioners to discuss the recent trends and challenges in building foundational models for human sensing, and integrating advances from Generative AI into sensor pipelines. The workshop will present an opportunity for researchers to present their latest findings, acquire hands-on experience, and discuss the future of the field. It will also contain a mini-tutorial and mini-hackathon on using Generative AI and foundational models for human sensing applications and host presentations for the accepted submissions, which will take the form of full and position papers. We particularly welcome preliminary or work-in-progress results, in cover topics such as, but not limited to:

• Foundational models for wearable and ubiquitous computing.
• Augmenting context descriptions with generative models.
• Multi-modal domain adaptation or representation Learning for wearable and ubiquitous computing.
• Using generative models for data simulation, e.g., text-to-motion, diffusion models, and others. This also includes physical simulation and addressing the critical "sim2real" gap.
• Large Language Model (LLM) integration for ubiquitous computing.
• Challenges in practical deployment of Generative AI and foundational models in-the-wild.

Papers submitted for review should be 4 - 6 pages in length (excluding references). Submissions will go through a single-phase review process with at least 2 reviewers. They will be reviewed and selected based on their originality, relevance, technical correctness, and their potential for initiating fruitful discussions at the workshop. Note, that position papers are not expected to present finished research projects. We rather ask for thought-provoking ideas or initial explorations of a topic. At the workshop, accepted submissions will be presented in a 10-minute presentation. At least one author of an accepted submission must attend the workshop. All accepted workshop papers can be included in the ACM Digital Library (DL) and the supplemental proceedings.