no code implementations • 29 Oct 2023 • Kyung Geun Kim, Byeong Tak Lee
Many diverse phenomena in nature often inherently encode both short- and long-term temporal dependencies, which especially result from the direction of the flow of time.
no code implementations • 24 Aug 2023 • Byeong Tak Lee, Yong-Yeon Jo, Joon-Myoung Kwon
Although ECG signals are time-series data, CNN-based models have been shown to outperform other neural networks with different architectures in ECG analysis.
no code implementations • 29 May 2022 • Kyung Geun Kim, Byeong Tak Lee
In this paper, we propose a novel graph-based data augmentation method that can generally be applied to medical waveform data with graph structures.