Jens Grubert (@armob) profile picture

Mixed Reality Researcher

Germany
Joined August 2009
Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)Photo by Jens Grubert (@armob)
Jens Grubert retweeted
Introducing Objaverse, a massive open dataset of text-paired 3D objects! Nearly 1 million annotated 3D objects to pave the way to build incredible large-scale 3D generative models: 🧵👇 🤗 Hugging Face: huggingface.co/datasets/alle… 📝ArXiv: arxiv.org/abs/2212.08051 #CVPR2023
248
482
331
2,638
1,601,605
Jens Grubert retweeted
Text2Tex is an exciting paper using Objaverse that also came out just this week by @davech2y et al. for texturing 3D meshes from a text description! nitter.cf/_akhaliq/status/163833…
Text2Tex: Text-driven Texture Synthesis via Diffusion Models abs: arxiv.org/abs/2303.11396 project page: daveredrum.github.io/Text2Te…
2
5
35
38,011
Jens Grubert retweeted
We’ve just launched the DALL·E API so developers can integrate DALL·E directly into their own apps and products. openai.com/blog/dall-e-api-n…
139
685
92
3,588
We then learn a generative model over latent representations using a diffusion model. This allows us to tackle both unconditional and conditional inference tasks. Like generating 3D scenes and camera trajectories from text prompts (additional results on github.com/apple/ml-gaudi):
1
4
3
30
Web3 und Metaverse: Experten warnen Bundestag vor "toxischer" Technik heise.de/news/Web3-und-Metav… #Web3 #Blockchain
4
14
20
Jens Grubert retweeted
Authentic Volumetric Avatars from a Phone Scan paper: dl.acm.org/doi/abs/10.1145/3…
4
85
7
456
Jens Grubert retweeted
How can robots perform a wide variety of novel tasks from natural language? Execited to present Code as Policies - using language models to directly write robot policy code from language instructions. See paper, colabs, blog, and demos at code-as-policies.github.io long 🧵👇
16
144
38
651
1/ From today's AI@ event: we announced our Imagen text-to-image model is coming soon to AI Test Kitchen. And for the 1st time, we shared an AI-generated super-resolution video using Phenaki to generate long, coherent videos from text prompts and Imagen Video to increase quality.
226
1,392
342
9,509
Jens Grubert retweeted
New open-source language model from Google AI: Flan-T5 🍮 Flan-T5 is instruction-finetuned on 1,800+ language tasks, leading to dramatically improved prompting and multi-step reasoning abilities. Public models: https://nitter.cf/t.co/bnYVnocJW2 Paper: arxiv.org/abs/2210.11416
37
460
69
2,335
One more thought: some examples are already of the usable quality level (better than some store asset packs) for indie, stylized/NPR games. Yes, it needs converting to a mesh, animating, etc. - all doable, “system” work and research. Absolute revolution in 3D content creation.🤯
Happy to announce DreamFusion, our new method for Text-to-3D! dreamfusion3d.github.io We optimize a NeRF from scratch using a pretrained text-to-image diffusion model. No 3D data needed! Joint work w/ the incredible team of @BenMildenhall @ajayj_ @jon_barron #dreamfusion
3
7
1
33
Want to use @StableDiffusion right from #Photoshop? Now you can! christiancantrell.com/#ai-ml
107
409
100
2,412
A year ago we open-sourced #AlphaFold Today we're freely sharing the predicted structures of all 200M+ proteins known to science - almost the entire protein universe! It's our gift to humanity, and a demonstration of the benefits AI can bring to society dpmd.ai/AF-22-LI
175
2,115
381
9,248
Jens Grubert retweeted
Our newest system DALL·E 2 can create realistic images and art from a description in natural language. See it here: openai.com/dall-e-2/
466
2,880
1,234
10,180
Jens Grubert retweeted
Very glad I can finally talk about our newly-minted #CVPR2022 paper. We extended mip-NeRF to handle unbounded "360" scenes, and it got us ~photorealistic renderings and beautiful depth maps. Explainer video: youtube.com/watch?v=zBSH-k9G… and paper: arxiv.org/abs/2111.12077
23
247
32
1,632
Every successful CHI author uses these 7 rules of writing. Most graduate students do not know them. I'm releasing them to you here for free to help you become a better writer. 🧵⬇️
4
37
5
144
Call for Journal Papers for IEEE ISMAR 2022! Selected papers will be published in a special issue of IEEE Transactions on Visualization and Computer Graphics (TVCG). #ISMAR22
Jens Grubert retweeted
New Year's Resolution: 5 µm The hyper-res image of Rembrandt's Night Watch @Rijksmuseum is now online: hyper-resolution.org/Nightwa… 97 rows × 87 cols (8 439) @Hasselblad H6D-400c MS images Size: 925 000 px × 775 000 px (717 GP) Res: 5 µm (5080 ppi)
8
69
37
209
Just released: The newest #AI100report examines the current state of AI, it's far-reaching impact, and the role of government, academia, the industry, and the public in shaping the technology. Read the report: stanford.io/3AguWBT
7
136
24
293
Jens Grubert retweeted
DINO’s attention maps can discover and segment objects in an image or a #Video with absolutely no supervision and without being given a segmentation-targeted objective. #computervision ai.facebook.com/blog/dino-pa…
Here’s our new computer vision system achieving state of the art results in image segmentation, without needing any labeled training data. This new model was trained on random, unlabeled data, but quickly achieved state-of-the-art results. It’s awesome.
4
70
2
282
#CHI2021 Online registration is open! Early registration is at a lower rate! Subject to Covid restrictions we plan a special gift for the first 1000 registrations. There are even lower rates for economically developing countries until Mar 16. Let's go! cvent.me/gRReaO
2
50
11
85