Albert Pumarola
@AlbertPumarola
Followers
1K
Following
3K
Media
33
Statuses
261
Researcher & Manager @Meta Superintelligence Labs
Zurich, Switzerland
Joined March 2014
Join us at the oral presentation and poster to learn more about Autoregressive Distillation. #CVPR2025.
Thrilled to share that our CVPR 2025 paper โ๐๐ฎ๐ญ๐จ๐ซ๐๐ ๐ซ๐๐ฌ๐ฌ๐ข๐ฏ๐ ๐๐ข๐ฌ๐ญ๐ข๐ฅ๐ฅ๐๐ญ๐ข๐จ๐ง ๐จ๐ ๐๐ข๐๐๐ฎ๐ฌ๐ข๐จ๐ง ๐๐ซ๐๐ง๐ฌ๐๐จ๐ซ๐ฆ๐๐ซ๐ฌโ(ARD) has been selected as an Oral! โจ. Catch us at CVPR on Saturday, June 14. ๐ฃ Oral Session 4A โ 14:00-14:15, Karl Dean Ballroom
0
0
9
We have new open positions:.Looking for strong candidates with experience on diffusion / flow matching models to work on the next generation image and video models in GenAI. Apply Now: ๐.
0
2
18
๐ 9ร Faster Llama!.Check out our latest #ICLR2025 introducing "Judge Decoding", a new simple paradigm that slashes LLM latency by 9ร. ๐Paper:
1
2
27
Best conference of the year. See you all there.
0
1
6
๐ขNow Hiring: PhD Interns at @Meta! . ๐ We're on the hunt for exceptional candidates to intern with our team and work on cutting-edge #DiffusionModels at Meta - GenAI. ๐ก. Apply Now:
2
23
152
We can finally share "Movie Gen" ๐ฅณ:.* 30B video model, 73K video context tokens. * T2V: 16s video generation at 1080p. * I2V: Video personalizaton. * V2V: Video text base editing. ๐ Checkout our paper with all the tricks:
1
19
106
RT @dlbcnai: Save the date: the Deep Learning Barcelona Symposium 2024 will be on Thursday 19th December. It will be held at La Salle - Uniโฆ.
0
11
0
Imagine Flash now powers high quality real-time image generation on and @WhatsApp (in the US). ๐ Paper: ๐ฅ๏ธ Try it (only US): ๐จโ๐ฆฑ Mark Zuck. talking about it: ๐บ Product ad: ๐.
It's here! Our new image generation is now faster and higher qualityโyou can create images from text in real-time as you type, animate images and turn them into GIFs. Imagine it. Create it. Have fun!. Now rolling out on WhatsApp and the Meta AI web experience in the US.
0
2
21
RT @AIatMeta: In addition to Llama 3, today weโre also publishing a new paper: Imagine Flash: Accelerating Emu Diffusion Models with Backwaโฆ.
0
189
0
๐ขIntroducing Imagine Flash .Buckle up! We're releasing a new model for real-time image generation. Available in and WA. ๐ Paper: . Amazing team: J.Kohler, @schoenfeldedgar, @artsiom_s,.@rsumbaly, P.Vajda, @alitabet @AIatMeta
6
9
63
๐ขInternship Opportunity Alert!. ๐ We're on the hunt for exceptional candidates to intern with our team and work on cutting-edge #DiffusionModels at @Meta GenAI. If you're a dedicated and passionate individual we'd love to have you on board. ๐ก.
17
20
161
๐ Exciting News!!. Hard work paid off! ๐ In #MetaConnect2023, MZ showcased our tech enabling high-res image generation in <5 seconds! ๐ผ๏ธ๐ฅ. Try it out on WhatsApp, Instagram, and Facebook. Huge thanks to our incredible team! @alitabet @artsiom_s @schoenfeldedgar and J. Kohler.
1
10
58
๐ข Seeking talented minds for Generative AI at Meta! ๐๐ง . We're looking for exceptional candidates to join our team and work on cutting-edge generative AI projects. We want to hear from you! ๐ฅ #AI #GenerativeAI #JobOpportunity. Apply here โก๏ธ
2
3
20
RT @dlbcnai: Today is Deep Learning Barcelona Symposium 2022 !. If you cannot make it to Campus Norr UPC, watch it on YouTube starting 9amโฆ.
0
17
0
This week in #NeurIPS2022 we will be presenting VisCo-Grids, a grid-based surface reconstruction method incorporating Viscosity and Coarea priors. Work developed at @MetaAI with @artsiom_s, @YarivLior, @alitabet and @lipmanya. ๐PDF:
2
17
96
This is definitely the most surreal tweet Iโve ever seen. Thank you for sharing @Michael_J_Black.
I entered "Estimating realistic 3D human avatars in clothing from a single image or video". In this case, it made up a fictitious paper and associated GitHub repo. The author is a real person (@AlbertPumarola) but the reference is bogus. (2/9)
0
1
29
For those looking to start working on neural fields this is an amazing piece of educational content ๐.
Neural fields are emerging as useful signal representations in computer vision & beyond. Our full-day introductory @CVPR tutorial on the topic is now public. Video: Slides: Web:
0
2
7
RT @cristiancanton: A great Q&A session on our @MetaAI recruiting event featuring Javier Romero, Roberto Dessรฌ (@robdessi) and Albert Pumarโฆ.
0
1
0
๐๐๐ a must-go event if you are in BCN, see you all there!!!.
The Deep Learning Barcelona Symposium will be back after the COVID-19 break. Save the dates of 22-23 December, 2021. Call for presenters open, submit your talk or poster of published work by October 31:. Spread the word.
0
1
5
We have just released H3D-Net :), our new method for high-quality 3D reconstruction of human heads. #ICCV2021 . ๐ฅProject: ๐PDF: ๐ฆDataset:
(1/5) Thrilled to present โH3D-Netโ, a method for reconstructing high-fidelity human heads in 3D from small sets of images. This is joint work with @GilTriginer, @jannaescur, @AlbertPumarola, @jaimestetix, @DocXavi, and @fmorenoguer. Project page:
1
1
24