A Hybrid Model using Fire Module for ECG-based Human Identification : Hybrid Model for ECG-based Human Identification

In the present study, a computation-cost effective DL model for ECG-based human identification is introduced. The study proposes a hybrid model of DenseNet with fire module to reduce its learnable parameters without affecting its performance. Experiments on a dataset of 300 users show encouraging results. The proposed DL model is suitable for ECG-based human identification/ authentication in limited capacity devices.