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

Recent progress in text-to-image (T2I) models enables high-quality imagegeneration with flexible textual control. To utilize the abundant visual priorsin the off-the-shelf T2I models, a series of methods try to invert an image toproper embedding that aligns with the semantic space of the T2I model. However,these image-to-text (I2T) inversion methods typically need multiple sourceimages containing the same concept or struggle with the imbalance betweenediting flexibility and visual fidelity. In this work, we point out that thecritical problem lies in the foreground-background entanglement when learningan intended concept, and propose a simple and effective baseline forsingle-image I2T inversion, named SingleInsert. SingleInsert adopts a two-stagescheme. In the first stage, we regulate the learned embedding to concentrate onthe foreground area without being associated with the irrelevant background. Inthe second stage, we finetune the T2I model for better visual resemblance anddevise a semantic loss to prevent the language drift problem. With the proposedtechniques, SingleInsert excels in single concept generation with high visualfidelity while allowing flexible editing. Additionally, SingleInsert canperform single-image novel view synthesis and multiple concepts compositionwithout requiring joint training. To facilitate evaluation, we design anediting prompt list and introduce a metric named Editing Success Rate (ESR) forquantitative assessment of editing flexibility. Our project page is:https://jarrentwu1031.github.io/SingleInsert-web/

SingleInsert: Inserting New Concepts from a Single Image into Text-to-Image Models for Flexible Editing