Poster
DTWNet: a Dynamic Time Warping Network
Xingyu Cai · Tingyang Xu · Jinfeng Yi · Junzhou Huang · Sanguthevar Rajasekaran

Thu Dec 12th 10:45 AM -- 12:45 PM @ East Exhibition Hall B + C #103

Dynamic Time Warping (DTW) is widely used as a similarity measure in various domains. Due to its invariance against warping in the time axis, DTW provides more meaningful discrepancy measurements between two signals than other dis- tance measures. In this paper, we propose a novel component in an artificial neural network. In contrast to the previous successful usage of DTW as a loss function, the proposed framework leverages DTW to obtain a better feature extraction. For the first time, the DTW loss is theoretically analyzed, and a stochastic backpropogation scheme is proposed to improve the accuracy and efficiency of the DTW learning. We also demonstrate that the proposed framework can be used as a data analysis tool to perform data decomposition.

Author Information

Xingyu Cai (University of Connecticut)
Tingyang Xu (Tencent AI Lab)
Jinfeng Yi (JD Research)
Junzhou Huang (University of Texas at Arlington / Tencent AI Lab)
Sanguthevar Rajasekaran (University of Connecticut)

More from the Same Authors