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 works have explored text-guided image editing using diffusion modelsand generated edited images based on text prompts. However, the models struggleto accurately locate the regions to be edited and faithfully perform preciseedits. In this work, we propose a framework termed InstructEdit that can dofine-grained editing based on user instructions. Our proposed framework hasthree components: language processor, segmenter, and image editor. The firstcomponent, the language processor, processes the user instruction using a largelanguage model. The goal of this processing is to parse the user instructionand output prompts for the segmenter and captions for the image editor. Weadopt ChatGPT and optionally BLIP2 for this step. The second component, thesegmenter, uses the segmentation prompt provided by the language processor. Weemploy a state-of-the-art segmentation framework Grounded Segment Anything toautomatically generate a high-quality mask based on the segmentation prompt.The third component, the image editor, uses the captions from the languageprocessor and the masks from the segmenter to compute the edited image. Weadopt Stable Diffusion and the mask-guided generation from DiffEdit for thispurpose. Experiments show that our method outperforms previous editing methodsin fine-grained editing applications where the input image contains a complexobject or multiple objects. We improve the mask quality over DiffEdit and thusimprove the quality of edited images. We also show that our framework canaccept multiple forms of user instructions as input. We provide the code athttps://github.com/QianWangX/InstructEdit.

InstructEdit: Improving Automatic Masks for Diffusion-based Image Editing With User Instructions