GPT-4VなどのマルチモーダルLLMで画像を分析する際、画像に「ドットマトリックス」を重ね合わせるだけで認識精度が大きく向上することが報告されています。
"Scaffolding Coordinates to Promote Vision-Language Coordination in Large Multi-Modal Models"
■研究の背景…
Contrastive RL provides a way to use contrastive learning methods to learn general-purpose goal-conditioned policies, uniting representation learning and RL. We recently got this working at scale with real robots! You can read more here: chongyi-zheng.github.io/stable_contras…
A short 🧵👇
A video of my recent talk on pre-training general-purpose robot models: youtu.be/Rn3devt8ZCg
Covers some recent work that paints a picture of what the future of multi-task, multi-robot, multi-domain generalist robotic models might look like.
ImageBind is the first AI model capable of binding information from six modalities. It equips machines with a holistic understanding that connects objects in photos with how they will sound, their 3D shape, temperature & how they move.
More ➡️ bit.ly/46hAJaY
HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Sampling
paper page: huggingface.co/papers/2301.02…
Volumetric scene representations enable photorealistic view synthesis for static scenes and form the basis of several existing 6-DoF video techniques. However, the volume…
Introducing the latest release from Stability AI: Breaking barriers with #SDXL 0.9!
SDXL 0.9 produces massively improved text-to-image and composition detail over the beta release and provides a leap in use cases for generative AI imagery. #StabilityAI
Unleash your creativity…
A new compact design for soft peristaltic pumps can be utilized to power soft #robots. Researchers at @wyssinstitute@Harvard used this approach to mix different fluids and make a pina colada.
Read more: scim.ag/35X
A paper I enjoyed reading recently: « The space of spaces (...) » KT Sturm. The definitive paper on the Gromov-Wasserstein geometric of metric measure spaces. The formula for the geodesics is amazing. arxiv.org/abs/1208.0434
Training Socially Aligned Language Models in Simulated Human Society
Social alignment in AI systems aims to ensure that these models behave according to established societal values. However, unlike humans, who derive consensus on value judgments through social interaction,…
82K Followers 842 FollowingEarn passive rewards with your router. 📡 Download the app on the App Store or Google Play and help build the future of decentralized connectivity!
22K Followers 1K FollowingBuilding AI that makes autonomous decisions using world models, artificial curiosity, and temporal abstraction @GoogleDeepMind
712K Followers 288 FollowingTogether with the AI community, we are pushing the boundaries of what’s possible through open science to create a more connected world.
4.3M Followers 3 FollowingOpenAI’s mission is to ensure that artificial general intelligence benefits all of humanity. We’re hiring: https://t.co/dJGr6Lg202
4K Followers 2K FollowingSoft Robotics is the #1 robotics journal, publishing world-class peer-reviewed research on the emerging technologies of soft and deformable robots.
3K Followers 535 FollowingReinforcement Learning (and Games)
- Core Team Researcher H company
- Full Professor on sabbatical
- Ex-Ubisoft / Ex-Facebook/ Ex-Criteo
No recent Favorites. New Favorites will appear here.