How do Foundation Models Compare to Skeleton-Based Approaches for Gesture Recognition in Human-Robot Interaction?
IEEE International Conference on Robot and Human Interactive Communication (RO-MAN), pp. 477
Abstract
Gestures enable non-verbal human-robot communication, especially in noisy environments like agile production. Traditional deep learning-based gesture recognition relies on task-specific architectures using images, videos, or skeletal pose estimates as input. Meanwhile, Vision Foundation Models (VFMs) and Vision Language Models (VLMs) with their strong generalization abilities offer potential to reduce system complexity by replacing dedicated task-specific modules. This study investigates adapting such models for dynamic, full-body gesture recognition, comparing V-JEPA (a state-of-the-art VFM), Gemini Flash 2.0 (a multimodal VLM), and HD-GCN (a top-performing skeleton-based approach). We introduce NUGGET, a dataset tailored for human-robot communication in intralogistics environments, to evaluate the different gesture recognition approaches. In our experiments, HD-GCN achieves best performance, but V-JEPA comes close with a simple, task-specific classification head—thus paving a possible way towards reducing system complexity, by using it as a shared multi-task model. In contrast, Gemini struggles to differentiate gestures based solely on textual descriptions in the zero-shot setting, highlighting the need of further research on suitable input representations for gestures.
Authors 6
-
Affiliation as printed
RWTH Aachen University,Chair for Computer Vision,Germany
-
Affiliation as printed
RWTH Aachen University,Chair for Computer Vision,Germany
-
Affiliation as printed
RWTH Aachen University,Chair for Computer Vision,Germany
-
Affiliation as printed
RWTH Aachen University,Chair for Computer Vision,Germany
-
Affiliation as printed
Corporate Research & Bosch Center for AI,Robert Bosch GmbH,Renningen and Hildesheim,Germany
-
Affiliation as printed
RWTH Aachen University,Chair for Computer Vision,Germany
Cited by 2 stored of 2
2 results
No patents citing this paper on Lens.org (checked 2026-10-06).