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

This work presents AnyDoor, a diffusion-based image generator with the powerto teleport target objects to new scenes at user-specified locations in aharmonious way. Instead of tuning parameters for each object, our model istrained only once and effortlessly generalizes to diverse object-scenecombinations at the inference stage. Such a challenging zero-shot settingrequires an adequate characterization of a certain object. To this end, wecomplement the commonly used identity feature with detail features, which arecarefully designed to maintain texture details yet allow versatile localvariations (e.g., lighting, orientation, posture, etc.), supporting the objectin favorably blending with different surroundings. We further propose to borrowknowledge from video datasets, where we can observe various forms (i.e., alongthe time axis) of a single object, leading to stronger model generalizabilityand robustness. Extensive experiments demonstrate the superiority of ourapproach over existing alternatives as well as its great potential inreal-world applications, such as virtual try-on and object moving. Project pageis https://damo-vilab.github.io/AnyDoor-Page/.

AnyDoor: Zero-shot Object-level Image Customization