New framework generates synthetic IMU data from video

Researchers have developed VSMP-IMU, a novel framework for generating synthetic Inertial Measurement Unit (IMU) data. This method uses video input to create semantically rich motion programs, which are then translated into virtual IMU signals. VSMP-IMU aims to overcome limitations in current human activity recognition datasets by providing controllable and wearable-domain-grounded synthetic data, improving performance in low-resource, class-imbalanced, and subject-generalization scenarios. AI

IMPACT This framework could significantly improve human activity recognition models by providing a scalable solution for generating labeled sensor data.

RANK_REASON The cluster contains a research paper detailing a new method for synthetic data generation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Lala Shakti Swarup Ray
• Semantic Motion Program
• VSMP-IMU

• paper
• product

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