kevinzhang25's profile picture. CS PhD @ @UofMaryland, computer vision,

prev: @AdobeResearch, @Google, @Berkeley_ai,

holding symbols lightly

Kevin Zhang

@kevinzhang25

CS PhD @ @UofMaryland, computer vision, prev: @AdobeResearch, @Google, @Berkeley_ai, holding symbols lightly

Kevin Zhang 已轉發

Introducing Ctrl-VI, a video sampling method allowing for a flexible set of user controls—ranging from coarse but easy-to-specify text prompts to precise camera/object trajectories. (1/n) arxiv.org/abs/2510.07670


Kevin Zhang 已轉發

what if you could combine diffusion models instantly? You would get exponentially better control (for free!!👀) This is exactly what we do. In ✨ coupled diffusion sampling ✨, diffusion models guide each other. The result? Diverse editing capabilities!


fast fantastic fashion fabrication

Bodies change and styles evolve. How can our clothing adapt with us? Refashion explores resizing and restyling garments from a set of reusable building blocks. Same parts, many possibilities. #UIST2025

rebeccayelin's tweet image. Bodies change and styles evolve. 
How can our clothing adapt with us?

Refashion explores resizing and restyling garments from a set of reusable building blocks. 

Same parts, many possibilities. 
#UIST2025


Kevin Zhang 已轉發

Every lens leaves a blur signature—a hidden fingerprint in every photo. In our new #TPAMI paper, we show how to learn it fast (5 mins of capture!) with Lens Blur Fields ✨ With it, we can tell apart ‘identical’ phones by their optics, deblur images, and render realistic blurs.

estheroate's tweet image. Every lens leaves a blur signature—a hidden fingerprint in every photo.

In our new #TPAMI paper, we show how to learn it fast (5 mins of capture!) with Lens Blur Fields ✨

With it, we can tell apart ‘identical’ phones by their optics, deblur images, and render realistic blurs.

Kevin Zhang 已轉發

Huge thanks to my amazing co-authors: @exfilmstudent , @rebeccayelin , Daniel Miau, Florian Kainz, Jiawen Chen, @ceciliazhang77, @DaveLindell & @kyroskutulakos 📄 Paper ➡️ blur-fields.github.io 💻 Code: coming soon! @ComputerSociety #ComputationalPhotography #IEEECS


Kevin Zhang 已轉發

Guardrails with custom polices are hard for models trained on safety and harm-related datasets. But what if you trained a guardian model on arbitrary rules? Introducing DynaGuard, a guardian model for custom policies: arxiv.org/abs/2509.02563

MonteBHoover's tweet image. Guardrails with custom polices are hard for models trained on safety and harm-related datasets. But what if you trained a guardian model on arbitrary rules?
Introducing DynaGuard, a guardian model for custom policies: arxiv.org/abs/2509.02563

one might think that tpot stands for that part of twitter, but it actually stands for that part of tpot


Kevin Zhang 已轉發

If you’re at SIGGRAPH 2025 in Vancouver, join us Thu 2 PM for our talk “Generative Neural Materials”! We introduce a universal neural material model for bidirectional texture functions and a complementary generative pipeline. 1/2


Kevin Zhang 已轉發

✨ Our paper Magic Fixup is accepted to ACM TOG! We show how dynamic videos can guide photo editing across many tasks — making this a solid baseline for future research. project page: magic-fixup.github.io paper: dl.acm.org/doi/10.1145/37…

HadiZayer's tweet image. ✨ Our paper Magic Fixup is accepted to ACM TOG!
We show how dynamic videos can guide photo editing across many tasks — making this a solid baseline for future research.

project page: magic-fixup.github.io
paper: dl.acm.org/doi/10.1145/37…

chatgpt sycophancy psychosis ❌ cursor + claude side project vibecoding flow state ✅


every time i type img[:3, ...] to chop off an alpha channel i'm like hmmm.... a familiar friend.... :3


Kevin Zhang 已轉發

There is 1 more week to submit non-archival extended abstracts to present at the Artificial Social Intelligence workshop @ICCVConference! We welcome work recently published in other venues (including the main ICCV conference) as well as works in progress!

Excited to announce the Artificial Social Intelligence Workshop @ ICCV 2025 @ICCVConference Join us in October to discuss the science of social intelligence and algorithms to advance socially-intelligent AI! Discussion will focus on reasoning, multimodality, and embodiment.

lmathur_'s tweet image. Excited to announce the Artificial Social Intelligence Workshop @ ICCV 2025 @ICCVConference

Join us in October to discuss the science of social intelligence and algorithms to advance socially-intelligent AI! Discussion will focus on reasoning, multimodality, and embodiment.


there needs to be stronger pushback against AI systems that will be only net harmful to humanity, like pika's "AI only social video app". we need to fight against the slop metaverse

it’s so over

khoomeik's tweet image. it’s so over


Kevin Zhang 已轉發

Some problems can’t be rushed—they can only be done step by step, no matter how many people or processors you throw at them. We’ve scaled AI by making everything bigger and more parallel: Our models are parallel. Our scaling is parallel. Our GPUs are parallel. But what if the…


Kevin Zhang 已轉發

Super hyped about this release!! Asimov has been a great coworker 🥰


Loading...

Something went wrong.


Something went wrong.