Roi Bar-On

[Roi Bar-On @roibar_on]

1/9
Excited to share EditP23! 🎨
Finally, a single tool for ALL your 3D editing needs:
✅ Pose & Geometry Changes
✅ Object Additions
✅ Global Style Transformations
✅ Local Modifications
All driven by one simple 2D image edit. It's mask-free ✨ and works in seconds ⚡️.
🧵

[Etai Sella @etai_sella]

Even today, with powerful image editing models, making fine-grained structural changes to 3D shapes remains a major challenge.

In our new #SIGGRAPH2026 paper, Prox-E, we use primitive-based abstraction to leverage VLMs for precise, reasoning-based 3D editing!

👇

[Gal Metzer @galmetz]

Excited to share our work accepted to #SIGGRAPH2026 ! Video generation models struggle with something few talk about: their transformations don't evolve smoothly. You get long boring stretches... then a sudden semantic jump where everything "catches up" at once.

[Shelly Golan @Shelly_Golan1]

1/7

When rewards conflict, what should RL post-training of diffusion models optimize?

In visual generation, objectives are often in tension:
Prompt adherence can conflict with source preservation.
Photorealism can conflict with stylization.

In our new paper, ParetoSlider, we introduce a multi-objective RL framework that trains a single diffusion model for continuous control over competing reward objectives 🧵

[Roi Bar-On @roibar_on]

Better graphics than GTA-VI will ever have? 👀 Video games are finally bridging the uncanny valley. Incredible work by @DanaCohenBar & team on RealMaster, using IC-LoRA to turn video diffusion models into a 2nd stage renderer that perfectly balances photorealism and fidelity 🔥

[Dana Cohen Bar @DanaCohenBar]

DLSS 5 is all over the timeline, and for good reason. In my internship at @AIatMeta we had the same idea: use a video model as a learned second-stage renderer on top of game engines. In our paper RealMaster, we make synthetic video look real while preserving scene fidelity 👇

[Sagi Polaczek 🦜 @PolaczekSagi]

[1/4] Sync about it… 💭✨
Editing a portrait video yet keeping it fully synced with the original across the entire sequence.
Read more about Sync-LoRA: https://t.co/4sAnR8FcZ1 🚀

[Shai Yehezkel @YehezkelShai]

Visual Diffusion Models are Geometric Solvers

We cast geometry as images: a plain diffusion model denoises into valid solutions. It is simple, general and effective.
Shown on Inscribed Square, Steiner Tree, and Maximum Area Polygonization - all classic hard problems.

[Roi Bar-On @roibar_on]

Fantastic paper by @nirgoren and the team! 📄✨
The core insight that initial noise is so correlated with the final image is wild 🤯.
This makes cryptographically-secure 🔒, model-free authorship verification a practical reality.

Awesome work! 👏

[Nir Goren @nirgoren]

The initial noise in diffusion models is surprisingly correlated with the final image.
Our NoisePrints paper exploits this to provide a lightweight, distortion-free, cryptographically secure watermark for proving authorship of generated images & videos, requiring no model access.

[Shelly Golan @Shelly_Golan1]

T2I models excel at realism, but true creativity means generating what doesn't exist yet. How do you prompt for something you can't describe? 🎨

We introduce VLM-Guided Adaptive Negative Prompting: inference time method that promotes creative image generation.

1/6

[Roi Bar-On @roibar_on]

Tired of complex prompts being ignored & static CFG limits? 😫

Huge congrats @YehezkelShai @OmerDahary @kusichan & @DanielCohenOr1 on their SIGGRAPH Asia paper! Their "Annealing Guidance" reacts as the image is created, adapting guidance for results a fixed value can't match. 🚀✨

Project: https://t.co/6Ml3IZLWKt

[Omer Dahary @OmerDahary]

Everyone uses CFG with w = 7.5… but why?? 🤔
For non-trivial prompts, choosing w is hard:
too low → weak alignment, too high → artifacts.
We show w shouldn’t be fixed — it should adapt! ✨
We present a tiny but smart MLP to adjust w over time
→ Better images/alignment 🚀

1/n

[Roi Bar-On @roibar_on]

@DanaCohenBar @DanielCohenOr1 For a complete overview of the project, including more examples and a video, please visit our project page: https://t.co/Fx4w7O8tZh
📷

[Roi Bar-On @roibar_on]

1/9
Excited to share EditP23! 🎨
Finally, a single tool for ALL your 3D editing needs:
✅ Pose & Geometry Changes
✅ Object Additions
✅ Global Style Transformations
✅ Local Modifications
All driven by one simple 2D image edit. It's mask-free ✨ and works in seconds ⚡️.
🧵

[Roi Bar-On @roibar_on]

9/9
Dive into the details or try it yourself!
📄 Paper: https://t.co/qPmwj4twpo
💻 Code: https://t.co/ECq1n1BsVK
🤗 Demo: https://t.co/LExzWJJvlq
Huge thanks to my amazing collaborators, @DanaCohenBar and @DanielCohenOr1 !

[Roi Bar-On @roibar_on]

@inbarhub @ICCVConference @logtdx @vd_kulikov @MatanKleiner @t_michaeli Totally earned 👏🔥