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 introduce BSD-GAN, a novel multi-branch and scale-disentangled trainingmethod which enables unconditional Generative Adversarial Networks (GANs) tolearn image representations at multiple scales, benefiting a wide range ofgeneration and editing tasks. The key feature of BSD-GAN is that it is trainedin multiple branches, progressively covering both the breadth and depth of thenetwork, as resolutions of the training images increase to reveal finer-scalefeatures. Specifically, each noise vector, as input to the generator network ofBSD-GAN, is deliberately split into several sub-vectors, each corresponding to,and is trained to learn, image representations at a particular scale. Duringtraining, we progressively "de-freeze" the sub-vectors, one at a time, as a newset of higher-resolution images is employed for training and more networklayers are added. A consequence of such an explicit sub-vector designation isthat we can directly manipulate and even combine latent (sub-vector) codeswhich model different feature scales.Extensive experiments demonstrate theeffectiveness of our training method in scale-disentangled learning of imagerepresentations and synthesis of novel image contents, without any extra labelsand without compromising quality of the synthesized high-resolution images. Wefurther demonstrate several image generation and manipulation applicationsenabled or improved by BSD-GAN. Source codes are available athttps://github.com/duxingren14/BSD-GAN.

BSD-GAN: Branched Generative Adversarial Network for Scale-Disentangled Representation Learning and Image Synthesis