ECG Beat Classification

This model incorporates both spatial as well as temporal feature extraction, and ensample classification. The model meaningfully reduces training cost as well as hardware needs by operating directly on raw 1D ECG signals as opposed to relying on signal-to-image conversions, which are computationally expensive, and is thus suitable in edge computing and wearable health and real-time applications.