Multiscale Manifold Warping
This addresses alignment challenges in applications such as bioinformatics and activity recognition, representing an incremental improvement over existing methods.
The paper tackles the problem of aligning temporal sequences with unequal dimensions by exploiting multiscale manifold latent structure, resulting in a new framework called Warping on Wavelets (WOW) that outperforms state-of-the-art methods like canonical time warping and manifold warping on real-world datasets.
Many real-world applications require aligning two temporal sequences, including bioinformatics, handwriting recognition, activity recognition, and human-robot coordination. Dynamic Time Warping (DTW) is a popular alignment method, but can fail on high-dimensional real-world data where the dimensions of aligned sequences are often unequal. In this paper, we show that exploiting the multiscale manifold latent structure of real-world data can yield improved alignment. We introduce a novel framework called Warping on Wavelets (WOW) that integrates DTW with a a multi-scale manifold learning framework called Diffusion Wavelets. We present a theoretical analysis of the WOW family of algorithms and show that it outperforms previous state of the art methods, such as canonical time warping (CTW) and manifold warping, on several real-world datasets.