A motion-language model fine-tuned on clinical biomechanics data answers structured questions about movement, scoring highly on activity recognition and gait speed estimation but weaker and underpowered on diagnosis and fall-history prediction.
Karen Liu, and Michael J
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
BiomechGPT: Extending Motion-Language Models to Clinical Motion Understanding
A motion-language model fine-tuned on clinical biomechanics data answers structured questions about movement, scoring highly on activity recognition and gait speed estimation but weaker and underpowered on diagnosis and fall-history prediction.