Krishna Kumar Singh

Research Scientist at Adobe Research

[Sumith Kulal @sumith1896]

Diffusion Circle happening today @CVPR #CVPR2026

Upper Lobby F at 3:30pm, come join us!

[Yiqing Liang @YiqingLiang2]

🎨 Got a paper on generative AI or creative content? Submit to the AI for Content Creation Workshop @CVPR2026!

⏰ Deadline: March 23 — only 7 days left!

📌 Topics include:
• Image, video & 3D generation
• Editing, stylization & domain transfer
• Multi-modal content creation
• Creative applications: VFX, fashion, graphic design, architecture & more

Why submit?
🔓 Non-archival & Flexible formats: 8-page full paper or 4-page extended abstract
🏆 Best student paper award
✈️ Travel grants available
💻 Can't make it to Denver? Join remotely via Zoom!

Don't let your work-in-progress miss this chance to get visibility at #CVPR2026.

🔗 https://t.co/S04pEPaGWN
Submission Website: CMT3 - https://t.co/yObkm8BRr4

[Xun Huang @xxunhuang]

You can easily combine pretrained LLM and DiT by fusing their attention layers.

[Richard Zhang @rzhang88]

Come check out our #CVPR posters tmrw (Thu) morning, 10-12:30!
- 4a: Ensembling Off-the-Shelf Models for GAN Training (*also an oral at 8:30*)
- 76a: Spatially-Adaptive Multilayer Selection for GAN Inversion & Editing
- 77a: On Aliased Resizing & Surprising Subtleties in GAN Eval

[AK @_akhaliq]

Spatially-Adaptive Multilayer Selection for GAN Inversion and Editing
abs: https://t.co/9tpvhXuaRw
project page: https://t.co/BmTZ8aRoEH

[Yong Jae Lee @yong_jae_lee]

@ak92501 Code for GIRAFFE HD has been released: https://t.co/vviYTPnhQ8

[AK @_akhaliq]

InsetGAN for Full-Body Image Generation
abs: https://t.co/wpdsbAn88f
project page: https://t.co/5zhDaJobvZ

[AK @_akhaliq]

GIRAFFE HD: A High-Resolution 3D-aware Generative Model
abs: https://t.co/9suIV5YxCn

[Adobe @Adobe]

#ProjectStrikeAPose leverages machine learning to recreate any pose on your photo subject, meaning you never have to worry where your hands go when taking a picture. #AdobeMAX https://t.co/Ye3558jvwj

[Ben Forta @benforta]

Our final #AdobeMAX2021 Sneak is #ProjectStrikeAPose. @Krishnakusin is creating brand new posed images from sample poses! And … no way! He took a selfie of his back and generated a matching view of the model from behind. But the source images were all front facing?! How the …?