Disentangled Dynamic Representations from Unordered Data
This work addresses the challenge of modeling sequential data without order constraints, which is incremental for applications like video analysis.
The authors tackled the problem of learning disentangled static and dynamic representations from unordered data, achieving a well-organized latent space for data dynamics as demonstrated on synthetic and real video datasets.
We present a deep generative model that learns disentangled static and dynamic representations of data from unordered input. Our approach exploits regularities in sequential data that exist regardless of the order in which the data is viewed. The result of our factorized graphical model is a well-organized and coherent latent space for data dynamics. We demonstrate our method on several synthetic dynamic datasets and real video data featuring various facial expressions and head poses.