HeartRate Estimation
The heartrate models detect person’s heartrate from frontal facial RGB video.
The training algorithm optimizes the network to minimize the root mean square error between predicted and ground truth pulse wave. This model was trained using the Heartrate Estimation training app in TLT v3.0.
The primary use case for this model is to detect heartrate. The model can be used to detect heartrate by using video with small compression ratio and proper frame rate.
The datasheet for the model is captured in its model card hosted at NGC.