U-Time: A Fully Convolutional Network for Time Series Segmentation Applied to Sleep Staging
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
U-Time : A Fully Convolutional Network for Time Series Segmentation Applied to Sleep Staging. / Perslev, Mathias; Hejselbak Jensen, Michael; Darkner, Sune; Jørgen Jennum, Poul; Igel, Christian.
Advances in Neural Information Processing Systems 32 (NIPS 2019). Vol. 32 NIPS Proceedings, 2019. p. 4415-4426.Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - U-Time
T2 - 33rd Conference on Neural Information Processing Systems (NeurIPS 2019)
AU - Perslev, Mathias
AU - Hejselbak Jensen, Michael
AU - Darkner, Sune
AU - Jørgen Jennum, Poul
AU - Igel, Christian
PY - 2019/10/1
Y1 - 2019/10/1
N2 - Neural networks are becoming more and more popular for the analysis of physiological time-series. The most successful deep learning systems in this domain combine convolutional and recurrent layers to extract useful features to model temporal relations. Unfortunately, these recurrent models are difficult to tune and optimize. In our experience, they often require task-specific modifications, which makes them challenging to use for non-experts. We propose U-Time, a fully feed-forward deep learning approach to physiological time series segmentation developed for the analysis of sleep data. U-Time is a temporal fully convolutional network based on the U-Net architecture that was originally proposed for image segmentation. U-Time maps sequential inputs of arbitrary length to sequences of class labels on a freely chosen temporal scale. This is done by implicitly classifying every individual time-point of the input signal and aggregating these classifications over fixed intervals to form the final predictions. We evaluated U-Time for sleep stage classification on a large collection of sleep electroencephalography (EEG) datasets. In all cases, we found that U-Time reaches or outperforms current state-of-the-art deep learning models while being much more robust in the training process and without requiring architecture or hyperparameter adaptation across tasks.
AB - Neural networks are becoming more and more popular for the analysis of physiological time-series. The most successful deep learning systems in this domain combine convolutional and recurrent layers to extract useful features to model temporal relations. Unfortunately, these recurrent models are difficult to tune and optimize. In our experience, they often require task-specific modifications, which makes them challenging to use for non-experts. We propose U-Time, a fully feed-forward deep learning approach to physiological time series segmentation developed for the analysis of sleep data. U-Time is a temporal fully convolutional network based on the U-Net architecture that was originally proposed for image segmentation. U-Time maps sequential inputs of arbitrary length to sequences of class labels on a freely chosen temporal scale. This is done by implicitly classifying every individual time-point of the input signal and aggregating these classifications over fixed intervals to form the final predictions. We evaluated U-Time for sleep stage classification on a large collection of sleep electroencephalography (EEG) datasets. In all cases, we found that U-Time reaches or outperforms current state-of-the-art deep learning models while being much more robust in the training process and without requiring architecture or hyperparameter adaptation across tasks.
KW - Computer Science - Machine Learning
KW - Electrical Engineering and Systems Science - Signal Processing
KW - Statistics - Machine Learning
M3 - Article in proceedings
VL - 32
SP - 4415
EP - 4426
BT - Advances in Neural Information Processing Systems 32 (NIPS 2019)
PB - NIPS Proceedings
Y2 - 8 December 2019 through 14 December 2019
ER -
ID: 239571874