LLM-Guided Exemplar Selection for Few-Shot Wearable-Sensor Human Activity Recognition
This addresses the need for efficient activity recognition in wearable devices with limited labeled data, though it is incremental by integrating LLM priors with existing geometric and structural techniques.
The paper tackled the problem of selecting representative exemplars for few-shot human activity recognition from wearable sensors, achieving an 88.78% macro F1-score on the UCI-HAR dataset, which outperformed classical methods like random sampling and herding.
In this paper, we propose an LLM-Guided Exemplar Selection framework to address a key limitation in state-of-the-art Human Activity Recognition (HAR) methods: their reliance on large labeled datasets and purely geometric exemplar selection, which often fail to distinguish similar wearable sensor activities such as walking, walking upstairs, and walking downstairs. Our method incorporates semantic reasoning via an LLM-generated knowledge prior that captures feature importance, inter-class confusability, and exemplar budget multipliers, and uses it to guide exemplar scoring and selection. These priors are combined with margin-based validation cues, PageRank centrality, hubness penalization, and facility-location optimization to obtain a compact and informative set of exemplars. Evaluated on the UCI-HAR dataset under strict few-shot conditions, the framework achieves a macro F1-score of 88.78%, outperforming classical approaches such as random sampling, herding, and k-center. The results show that LLM-derived semantic priors, when integrated with structural and geometric cues, provide a stronger foundation for selecting representative sensor exemplars in few-shot wearable-sensor HAR.