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

Existing person video generation methods either lack the flexibility incontrolling both the appearance and motion, or fail to preserve detailedappearance and temporal consistency. In this paper, we tackle the problem ofmotion transfer for generating person videos, which provides controls on boththe appearance and the motion. Specifically, we transfer the motion of oneperson in a target video to another person in a source video, while preservingthe appearance of the source person. Besides only relying on one source frameas the existing state-of-the-art methods, our proposed method integratesinformation from multiple source frames based on a spatio-temporal attentionmechanism to preserve rich appearance details. In addition to a spatialdiscriminator employed for encouraging the frame-level fidelity, a multi-rangetemporal discriminator is adopted to enforce the generated video to resembletemporal dynamics of a real video in various time ranges. A challengingreal-world dataset, which contains about 500 dancing video clips with complexand unpredictable motions, is collected for the training and testing. Extensiveexperiments show that the proposed method can produce more photo-realistic andtemporally consistent person videos than previous methods. As our methoddecomposes the syntheses of the foreground and background into two branches, aflexible background substitution application can also be achieved.

Multi-Frame Content Integration with a Spatio-Temporal Attention Mechanism for Person Video Motion Transfer