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 focus on controllable disentangled representation learning (C-Dis-RL),where users can control the partition of the disentangled latent space tofactorize dataset attributes (concepts) for downstream tasks. Two generalproblems remain under-explored in current methods: (1) They lack comprehensivedisentanglement constraints, especially missing the minimization of mutualinformation between different attributes across latent and observation domains.(2) They lack convexity constraints, which is important for meaningfullymanipulating specific attributes for downstream tasks. To encourage bothcomprehensive C-Dis-RL and convexity simultaneously, we propose a simple yetefficient method: Controllable Interpolation Regularization (CIR), whichcreates a positive loop where disentanglement and convexity can help eachother. Specifically, we conduct controlled interpolation in latent space duringtraining, and we reuse the encoder to help form a 'perfect disentanglement'regularization. In that case, (a) disentanglement loss implicitly enlarges thepotential understandable distribution to encourage convexity; (b) convexity canin turn improve robust and precise disentanglement. CIR is a general module andwe merge CIR with three different algorithms: ELEGANT, I2I-Dis, and GZS-Net toshow the compatibility and effectiveness. Qualitative and quantitativeexperiments show improvement in C-Dis-RL and latent convexity by CIR. Thisfurther improves downstream tasks: controllable image synthesis, cross-modalityimage translation, and zero-shot synthesis.

Encouraging Disentangled and Convex Representation with Controllable Interpolation Regularization