
This paper is both a nice tip of the hat to the films of Danny Boyle, and another step in our ongoing quest to reconstruct everything in the world in 3D. Check out more at the webpage here! hanyuc.com/trainspotting/

3D vision fanatic. Professor @cornell_tech & Researcher @GoogleDeepmind. He or they. snavely.bsky.social

This paper is both a nice tip of the hat to the films of Danny Boyle, and another step in our ongoing quest to reconstruct everything in the world in 3D. Check out more at the webpage here! hanyuc.com/trainspotting/

And here's the LNER Class A3 4472 “Flying Scotsman” locomotive used in the teaser: hanyuc.com/trainspotting/exp…

I love browsing Hanyu's interative dataset explorer -- here's a nice reconstruction of a Leica M3: hanyuc.com/trainspotting/exp…

Choose life. Choose a train, a statue, a pipe organ. Choose a 1967 Volkswagen Beetle, a K6 red telephone box in London. We chose not to choose. We reconstructed them all. Introducing 🚂 📸 Trainspotting: 12,068 objects reconstructed in 3D from 311,886 found photos. 🧵 1/8

Check out this cool work from @hanyuc1110 on creating really useful new data for training 3D foundation models like VGGT! He uses the wonderful zoo of images on Wikimedia Commons to reconstruct very nice 3D models of objects from vintage Atari computers to Ford Fiestas.
Choose life. Choose a train, a statue, a pipe organ. Choose a 1967 Volkswagen Beetle, a K6 red telephone box in London. We chose not to choose. We reconstructed them all. Introducing 🚂 📸 Trainspotting: 12,068 objects reconstructed in 3D from 311,886 found photos. 🧵 1/8

Please apply to come be a postdoc at Cornell through this wonderful Empire AI Fellows program! I'll be looking to recruit a postdoc myself alongside the rest of the vision group at Cornell Tech in NYC.
Cornell (and me) is hiring postdocs, with focus on technical AI/ML research. Deadline for full consideration is Dec 15. I am recruiting through this call as well. I encourage applying as early as possible, because I am starting to review folders now! Application link ⬇️

Lucky to have a signed copy!
Honored and humbled to receive the ECVA PhD Award at #ECCV2026 @eccvconf! So grateful to my advisor @mapo1 for his guidance throughout, to all my brilliant coauthors over the years, and to the computer vision community for the support and the journey together. Thank you all

🚀 We are excited to release the dataset, code, and expanded paper for ORBIT++: Benchmarking SfM in the Wild with 360° Video! Following our initial publication at CVPR, we’re releasing the full benchmark and codebase: 🌐 Project: orbit-sfm.github.io/ 📄 arXiv: arxiv.org/abs/2608.22039 💻 Code/Data: github.com/google-deepmind/o… (1/3)

Introducing FilmGPT 🎬 (#SIGGRAPH2026) Can we learn the grammar of film directly from movies? We train an AR transformer on feature films to edit raw footage into coherent cinematic sequences, without generating new pixels. Project page: mudtriangle.com/filmgpt Demo below👇

Picture stepping into a live 3D broadcast or 3D video call and looking around freely. That is the promise of dynamic NVS. But multi-view camera setups are fixed so parts stay hidden. Looking anywhere means recovering them from memory, which current stateful and stateless models cannot do in real-time over long-horizons. Until now. Introducing NSTM. ⚡️Real-time inference + 🧠minute-long memory.

We use a sort of @elliottszwu-style strategy to build a dataset of symmetries in architectural scenes, then detect and localize symmetries in new images via signed-distance functions. This is really great work from Hanyu with very nice visuals!
Excited to share ArchSym at #CVPR2026! 🏛️ Existing 3D symmetry detectors work well on clean, object-centric data. But what about in the wild? In our work, we tackle 3D-grounded reflectional symmetry detection specifically for real-world architectural landmarks. 🧵[1/7]

Thank you, Humphrey! It's a real honor to receive this award!
Congratulations to Noah Snavely @Jimantha from @Cornell receiving the 2026 Thomas S. Huang Memorial Prize at @CVPR — a well-deserved recognition of his impactful contributions to computer vision research and our community. Remembering Tom & Margaret: youtu.be/QV7WnO9Lk9M?si=iYmA…

Excited to share ArchSym at #CVPR2026! 🏛️ Existing 3D symmetry detectors work well on clean, object-centric data. But what about in the wild? In our work, we tackle 3D-grounded reflectional symmetry detection specifically for real-world architectural landmarks. 🧵[1/7]

Honey, I Shrunk the Arc de Triomphe! 😱 Ever notice how SOTA depth models suffer from "scale-collapse"—metrically shrinking distant landmarks like they're toys? We introduce MetricScenes: a new in-the-wild metric dataset that fixes this!

A really nice paper with a really cool title from Yuanbo, Hanyu, and Xueqing!
Honey, I Shrunk the Arc de Triomphe! 😱 Ever notice how SOTA depth models suffer from "scale-collapse"—metrically shrinking distant landmarks like they're toys? We introduce MetricScenes: a new in-the-wild metric dataset that fixes this!

This work has been fully open-sourced, including all code, model weights, and data related scripts. Feel free to check it out! Our CVPR poster session will be on Saturday, June 6, from 4:45 to 6:45 PM in ExHall A & F 2. Our team (@holynski_, @jon_barron , @ChrisWu6080, @Jimantha, and me) would be onsite and very happy to discuss it in person!
Spatial reconstruction is a long-context problem: real scenes come with hundreds of images. But O(N²) transformer-based models don’t scale efficiently. Introducing: 🤐ZipMap (CVPR ’26): Linear-Time, Stateful 3D Reconstruction via Test-Time Training (TTT). ZipMap “zips” a large image collection into an implicit TTT scene state in a single linear-time operation. The state will then be decoded into spatial outputs, and can be queried efficiently for novel-view geometry and appearance (~100 FPS) ZipMap is not only much faster (>20× faster than VGGT), but also matches or surpasses the accuracy of all SOTA models.