Fusion-E2Pulse: A Multimodal Event-RGB Fusion Network for Non-contact Pulse Wave Reconstruction
Researchers have developed a new multimodal network called Fusion-E2Pulse to improve non-contact pulse wave reconstruction. This system combines traditional RGB video data with signals from neuromorphic event cameras. The RGB data helps to reduce motion artifacts, while the event camera data captures fine-grained details of vascular pulsations. This fusion approach aims to overcome the limitations of each individual method, leading to more accurate waveform morphology recovery. AI
IMPACT This research introduces a novel AI approach for enhanced physiological signal extraction, potentially improving remote health monitoring.