Multi-modal sensing
Shen AI uses video input from a camera to detect the face and extract physiological vital signs. This is done by analyzing subtle pulsation signals from the skin, captured through remote photoplethysmography (rPPG). However, factors like lighting, skin tone, and movement can affect signal qualityāmaking consistent accuracy a challenge.
ā
To address this, Shen AI integrates multimodal sensing, combining rPPG with remote ballistocardiography (rBCG). This approach enhances accuracy by fusing signals: when rPPG is limited, rBCG helps maintain reliable measurements. Multimodal sensing also ensures that Shen AI works across a wide range of lighting conditions and skin tonesāmaking device-free health monitoring more accurate, inclusive, and accessible.