English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We construct an unsupervised learning model that achieves nonlineardisentanglement of underlying factors of variation in naturalistic videos.Previous work suggests that representations can be disentangled if all but afew factors in the environment stay constant at any point in time. As a result,algorithms proposed for this problem have only been tested on carefullyconstructed datasets with this exact property, leaving it unclear whether theywill transfer to natural scenes. Here we provide evidence that objects insegmented natural movies undergo transitions that are typically small inmagnitude with occasional large jumps, which is characteristic of a temporallysparse distribution. We leverage this finding and present SlowVAE, a model forunsupervised representation learning that uses a sparse prior on temporallyadjacent observations to disentangle generative factors without any assumptionson the number of changing factors. We provide a proof of identifiability andshow that the model reliably learns disentangled representations on severalestablished benchmark datasets, often surpassing the current state-of-the-art.We additionally demonstrate transferability towards video datasets with naturaldynamics, Natural Sprites and KITTI Masks, which we contribute as benchmarksfor guiding disentanglement research towards more natural data domains.

Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding