180 ActiLLM in progress

Research on activity recognition typically requires the processing of data from extensive sensor arrays and the development of detection models tailored to each specific usage context. These requirements limit the reuse potential of models and often lead to a duplication of research efforts. The recent surge in popularity of Large Language Models (LLMs) has underscored that these pre-trained models possess the ability to be applied across a wide range of applications with minimal need for adjustments. This mandate will focus on evaluating LLMs ability to accurately identifying various activities taking place within a smart home environment.

Lead applicant