Dynamic Variational Autoencoders for Visual Process Modeling

Alexander Sagel, Hao Shen

This work studies the problem of modeling visual processes by leveraging deep generative architectures for learning linear, Gaussian representations from observed sequences. We pro…