Adobe Research and the InDesign team joined forces for the Summit Sneak #ProjectVisionCast, an experimental prototype that allows users to combine data insights and explore brand imagery, fostering creative brainstorming backed by numbers. research.adobe.com/news/the-resea…

Today @ ExHall D Poster #176 We will present UniReal, our multimodal image generation and editing model. Come and drop by if you are interested #CVPR2025 Adobe Research

Yang Hansen was sitting in the stands when he got drafted Basketball guy

DiffusionGS is accepted in ICCV'2025 where we bake gaussion spaltting into the diffusion model so that we don't need to do multiple view generation and reconstruction as two separate steps. Adobe Research

Nice work!! We are trying almost exactly same idea to unified visual representation for both understanding and generation

We presented OmniVcus, a unified video customization framework can do (1) Single/Double/Multi-subject Video Cus Video Cus (2) Instructive Edit Subject-driven Video Cus (3) Camera-controlled Subject-driven Video Cus (4) Depth/mask -controlled Subject-driven Video Customization

Photoshop’s new harmonize feature looks genuinely useful — effectively making complex compositing tasks just one click. Seems Adobe has productized Project Perfect Blend from their sneaks presentation.

It seems that people like our feature Mengwei Ren

First unveiled as early research at #AdobeMAX last year, Adobe Photoshop’s new Harmonize feature can turn hours of editing into minutes! Kudos to the Adobe Research and Photoshop Engineering teams! adobe.ly/4lxiXXs

Nice work ! Glad to see more research on treating image editing as a video generation task like we did with Xi Chen on UniReal (CVPR'25) arxiv.org/abs/2412.07774

I will be at #ICCV2025 presenting two papers in the main conference! 🚀 Details are in the thread below. In the meantime, we’re hiring interns and full-time research scientists passionate about generative modeling — feel free to reach out if you’d like to join us or chat at ICCV!

The first one is DiffusionGS, a single-stage diffusion model that bridges 2D generation and 3D Gaussian splatting for fast, high-quality 3D scene reconstruction from just one image. caiyuanhao1998.github.io/project/Diffus…

And the 2nd one: DIVE: Taming DINO for Subject-Driven Video Editing, which leverages powerful DINO features to achieve consistent, identity-preserving video edits guided by text or reference images.