New KD-Judge framework uses LLMs to automate fitness movement assessment

Researchers have developed KD-Judge, a new framework designed to automatically and accurately assess functional fitness movements. This system converts unstructured rulebook standards into machine-readable formats using LLM-based retrieval and chain-of-thought reasoning. KD-Judge then employs a deterministic rule-based judging system with pose-guided kinematic analysis to validate repetitions and temporal boundaries, offering transparency and efficiency. The framework is optimized for edge devices like the Jetson AGX Xavier, incorporating a dual strategy caching mechanism to reduce computational load and achieve significant speedups. AI

IMPACT Enables more objective and efficient evaluation of physical performance, potentially impacting training, competition, and health monitoring.

RANK_REASON Research paper detailing a new framework for automated assessment of functional fitness movements. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CV →

• alphaXiv
• arXiv
• CatalyzeX
• CFRep
• DagsHub
• Gotit.pub
• Hugging Face
• Jetson AGX Xavier
• KD-Judge
• Shaibal Saha

• paper
• product
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →