On Disentanglement in Gaussian Process Variational Autoencoders
This work addresses the challenge of disentangling independent factors in sequential data, which is incremental as it builds on existing models by explicitly exploiting temporal structure.
The paper tackled the problem of learning disentangled representations from complex multivariate time series by investigating Gaussian process variational autoencoders, demonstrating competitiveness against state-of-the-art methods on a benchmark task and showing meaningful results on real-world medical data.
Complex multivariate time series arise in many fields, ranging from computer vision to robotics or medicine. Often we are interested in the independent underlying factors that give rise to the high-dimensional data we are observing. While many models have been introduced to learn such disentangled representations, only few attempt to explicitly exploit the structure of sequential data. We investigate the disentanglement properties of Gaussian process variational autoencoders, a class of models recently introduced that have been successful in different tasks on time series data. Our model exploits the temporal structure of the data by modeling each latent channel with a GP prior and employing a structured variational distribution that can capture dependencies in time. We demonstrate the competitiveness of our approach against state-of-the-art unsupervised and weakly-supervised disentanglement methods on a benchmark task. Moreover, we provide evidence that we can learn meaningful disentangled representations on real-world medical time series data.