"Diffusion Illusions: Hiding Images in Plain Sight": In this fun interactive demo, we use text-to-image models to create optical illusions and hide pictures inside each other! Visit ryanndagreat.github.io/Diffusion-Illu… to try it out! Source code included.

In collaboration with Baiyu Shi @ayah_ahmadd Roy Lin Ken Goldberg UC Berkeley + Daniel Seita 🇺🇦 David Held CMU Robotics Institute + Richard Cheng Thomas Kollar Toyota Research Institute (TRI) We thank Ryan Burgert for providing the cute rubber ducks! Check out our paper! (11/11) arxiv.org/abs/2303.16898

Why won't GPT4 speak to me in english? Haalp!

Our award-winning CVPR demo Diffusion Illusions now has a paper and even more examples! Please check it out at diffusionillusions.com

this project generates optical illusions and performs stenography using off-the-shelf diffusion models - all of which work in the physical world too! Paper: Diffusion Illusions: Hiding Images in Plain Sight by Ryan Burgert and team Link: arxiv.org/abs/2312.03817 Project:

#NeurIPS2023 🔔 Wondering an agent can learn how to see 👀 to help act 🦾? Come to see our #ActiveVision #RL with great potential! Time: Thursday, Dec 14 10:45 - 12:45 CST Venue: Great Hall & Hall B1+B2 (level 1) #1501 Everything: elicassion.github.io/sugarl/sugarl.…

🚀 Excited to share our latest project: LLaRA - Supercharging Robot Learning Data for Vision-Language Policy! 🤖✨ We create a framework to turn robot expert trajectories into conversation-style data and other auxiliary data for instruction tuning. More details to come! (1/N)

Want to know visual instruction tuning for robotics? 🤖Check our latest work -- LLaRA

I'm on this! :D

Thrilled to present 💡DifFRelight💡 at #SIGGRAPHAsia2024, a novel-view facial dynamic relighting framework based on image diffusion editing, dynamic Gaussian Splatting conditioning, integrated lighting control, and our in-house light stage hardware. Kudos to the team: Mingming

Our team has arrived in Munich and we're thrilled to present this work at the LangRob Workshop @ #CoRL2024 as a spotlight presentation on Nov. 9 morning. Stay tuned!

Kudos to the research team at our sister company Eyeline Studios. Their latest research paper, 🌊Go-with-the-Flow 🌊, will be presented at #CVPR2025! Based on their research, we believe this could allow artists in the future to leverage these new techniques to direct the motion

The first project I led at Netflix Eyeline Studios is headed to #CVPR2025 with 5,5,4 review scores: 🌊Go-with-the-Flow🌊 warps noise for effortless motion control in video diffusion — no pipeline changes, same compute. Direct camera/object motion, transfer movement between