Torsten Sattler

@CSProfKGD Those were the days :) Enjoy Zürich!

In Zurich is where I first met @SattlerTorsten ❤️ These trips are valuable networking experiences!

See you soon 🤗

@gabriberton In the end, the whole AGI debate is, in my opinion, a bit pointless: we don't need AGI for AI to be useful, the same way that robots have been useful for decades without ever (needing to) reach full human capabilities.

@SattlerTorsten Lol

Prediction: in 10 years we might have AGI but we'll still be using backprop

@Parskatt @gabriberton I predict that in 10 years, it will definitely be April 2036 ;)

@gabriberton @SattlerTorsten In 10 years I predict it likely to be 2036.

@gabriberton That is why I feel confident in my prediction: I can always move the goalpost in terms of what AGI means :) . Although it seems that the requirements for AGI are going down rather than up, so likely someone will claim to have achieved AGI for a rather narrow definition of AGI :)

@SattlerTorsten That will very much depend on the definition of AGI, according to some AGI is already here (LLMs pass the Turing test which is pretty cool)

@Michael_J_Black Congratulations!

Tomorrow, I start a new full-time position as VP Digital Human Research at Epic Games. (And no, this isn’t an April Fools’ joke.) Today, I’m taking early retirement from Max Planck and will become an Emeritus Director. As an Emeritus Director, I will continue to supervise my remaining students, oversee ongoing projects, and wind down my department over the next couple of years, which is a normal process when a director retires. Being a Max Planck director is the best academic job in the world and it has been my great honor to co-found and help build the Max Planck Institute for Intelligent Systems. I love the institute deeply. But my statutory retirement age was looming and I’m not done yet. I’ve been working on capturing and modeling human movement for 30 years and it has gone from a fringe topic to something that works robustly and has immense potential. With AI today, scale matters, and achieving that scale increasingly requires industry. In particular, digital humans are moving from research prototypes to foundational technology across industries. At this point in my career, I want to get this technology into the hands of millions of users, while pushing the frontier of digital humans. I know that this is a challenging time for the games industry. Such times are precisely when people get creative, the industry is open to change, and real innovation can take root. Epic Games is the ideal place for this. The talent is deep, there is a compelling vision for how games will evolve, and the commitment is clear. I’m excited to join them and the rest of the Meshcapade team. There is no good way to express on social media the depth of my gratitude to the Max Planck Society, my co-directors, students, post docs, and staff. It has been an amazing 15-year journey because of you. Thank you.

@janusch_patas Thanks for reposting :)

The Role and Relationship of Initialization and Densification in 3D Gaussian Splatting Contributions: (i) Even if accurate and dense point positions are available, densification can still improve NVS quality. Furthermore, for certain densification strategies, using dense initialization can actually degrade performance. (ii) Despite this, using dense initialization can improve generalization performance on views underrepresented in the training set. (iii) Two of the evaluated densification strategies [9, 21] are highly insensitive to both the accuracy and density distribution of the initial set of Gaussians. Thus, with the current densification methods, initialization based on stereo matching and monocular depth can be used in practice to improve generalization, without requiring significant additional processing time. Generally, we observe that there is no densification strategy that performs best in all scenarios. (iv) Most importantly, we see little benefit in using dense initialization compared to sparse SfM point clouds.

Just as we say goodbye to Vancouver (YVR)🇨🇦🥹, it’s time to reveal our next destination! ✈️ From the Pacific coast to the Mediterranean sea... We are thrilled to announce that #3DV2027 is heading to Thessaloniki, Greece (SKG)! 🇬🇷🥳 See you there in Mid-April 2027! 🌍

And finally, the 3DV Outstanding Doctoral Dissertation Award goes to Michael Niemeyer! Congratulations! @Mi_Niemeyer Thesis title: "Neural Scene Representations for 3D Reconstruction and Generative Modeling" #3DV2026

3DV Outstanding Doctoral Dissertation Award Honorable Mention goes to Songyou Peng! @songyoupeng Thesis title: "Neural Scene Representations for 3D Reconstruction and Scene Understanding" #3DV2026

@gabriberton @abursuc Actually, ECCV outranks ICCV by now ;) scholar.google.com/citations?view…

@abursuc This x.com/gabriberton/st…

Scrolling back and forth through submissions in that infamous single-column 14-page template this week.

@holynski_ I was thinking of writing a paper "Building Rome in a Cloudy Afternoon" today ;)

also, Building Rome in a Day was like 17 years ago feels like it might be time to try building rome in just a few seconds? could be a fun follow up. who's game? one more internship slot available for this summer. DM me.

@gabriberton Because it is extremely fast, doesn't require a GPU, and works well for frame to frame matching?

How come ORB is the 4th most downloaded matcher in VisMatch ?!? Yes ORB from 2011 2011 is not a typo

@ZGojcic @Parskatt @Eng_Hemdi Congratulations on the CVPR paper! Very nice work. I don't see anyone accusing anyone else of plagiarism though (disclaimer: I didn't read the papers), just someone pointing out relevant concurrent work (given the amount of works published in vision, it's easy to miss works).

@Parskatt @Eng_Hemdi Even if you choose to raise the issue publicly, it would still have been nice to give the authors a bit more than an hour to respond to your comment before publicly accusing them of plagiarism.

We're releasing DiffusionHarmonizer, an online diffusion enhancer bridging neural reconstruction and photorealistic simulation by correcting artifacts, and harmonizing inserted objects so they truly belong in the scene: matching shadows, lighting & color research.nvidia.com/labs/sil/proje…

@Michael_J_Black @meshcapade Congratulations!

Big news: @Meshcapade is now part of Epic Games. This is a perfect match for our technology and team and I am super excited about what we will build together. I want to thank our many supporters who have provided funding and/or advice on our journey from 2018 to today, including @MP_Innovation, @maxplanckpress, @MPI_IS, @PerceivingSys, @matrixvc, @dcstalder, @dianaberlin, @HV_Capital, Zuzanna Czapinska, LBBW VC, @GoodwaterCap, CLO Virtual Fashion (@itsclo3d), @grbradsk, @lucvincent, @JeffDean, @ballmatthew, Andrew Hamel, Bill O’Farrell, Ammar Zakiullah, @Nicolas_Keller, Alex Diehl, @goodwinlaw, YPOG, @NVIDIA Inception Program, @msft4startups, and many more! Most importantly, I want to thank my co-founders @naureenmahmood and Talha Zaman and the whole @Meshcapade team. There is nothing more rewarding than working with great people who you like and trust to build products that customers love using technology you believe in. Thank you, thank you, thank you. Now, on to the next phase! mpg.de/26082348/max-p…

@ducha_aiki Fun fact: Switzerland stores energy by pumping water up mountains (e.g., through potential energy).

TIL about Energy Vault: revolutionary startup, who proposes to store energy by lifting stones up. en.wikipedia.org/wiki/Energy_Va… P.S. Not so surprising, after the fact that boiling water is the way to use nuclear energy.

@ducha_aiki Interesting that they are not even citing our Patch2Pix paper.

XRefine: Attention-Guided Keypoint Match Refinement Jan Fabian Schmid, Annika Hagemann tl;dr: pairwise kpt2subpix with self-attention and study how kpt accuracy influences the camera pose accuracy arxiv.org/abs/2601.12530

Call for WiCV Workshop Organizers at ECCV 2026! Interested in organizing the Women in Computer Vision (#WiCV) workshop at #ECCV2026? Apply by Jan 06, 2026: forms.gle/rdKbyTB2VRWK2D… Please Retweet and share! We look forward to your application.

@CSProfKGD Probably a healthy decision :)

@SattlerTorsten Not something I will indulge in regularly 😉

When in Germany … do as the Germans 😅

@CSProfKGD Probably "when in southern Germany" :) The region of large schnitzels certainly extends further, e.g., definitely to Zürich. I'd assume you can find them in Prague as well.

@SattlerTorsten When in Bavaria then? 🤓