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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

We study the problem of synthesizing a long-term dynamic video from only asingle image. This is challenging since it requires consistent visual contentmovements given large camera motions. Existing methods either hallucinateinconsistent perpetual views or struggle with long camera trajectories. Toaddress these issues, it is essential to estimate the underlying 4D (including3D geometry and scene motion) and fill in the occluded regions. To this end, wepresent Make-It-4D, a novel method that can generate a consistent long-termdynamic video from a single image. On the one hand, we utilize layered depthimages (LDIs) to represent a scene, and they are then unprojected to form afeature point cloud. To animate the visual content, the feature point cloud isdisplaced based on the scene flow derived from motion estimation and thecorresponding camera pose. Such 4D representation enables our method tomaintain the global consistency of the generated dynamic video. On the otherhand, we fill in the occluded regions by using a pretrained diffusion model toinpaint and outpaint the input image. This enables our method to work underlarge camera motions. Benefiting from our design, our method can betraining-free which saves a significant amount of training time. Experimentalresults demonstrate the effectiveness of our approach, which showcasescompelling rendering results.

Make-It-4D: Synthesizing a Consistent Long-Term Dynamic Scene Video from a Single Image