Image created with gemini-3.1-flash-image-preview with claude-opus-4.7. Image prompt: A Byzantine gold-ground mosaic icon of a stylized flattened saint cradling a hammered-gold armillary sphere containing a tiny tesserae landscape with mosaic mountains, rivers, and a spiral of clouds, halo rendered as an engraved circuit-ring, candlelit sacral glow on burnished gold with deep imperial purple and Tyrian crimson robes, symmetrical centered composition with generous negative space, bold ivory Trajan Roman capitals reading ‘WORLD MODELS’ across the lower third, painterly tactile grain, 16:9.
Google just turned Street View into a video game. The mother lode of ground level data — 280 billion real world panoramas, now playable in real time. Here’s everything you need to know in 7 mins: 00:00 Genie 3 Grounded In Reality 00:44 Real-time Demos! 03:48 The Bigger
https://x.com/bilawalsidhu/status/2057262850209419553
Project Genie 🤝 @GoogleMaps Street View You can now take real U.S. places and transform them into new, interactive worlds. 🌍
https://x.com/GoogleDeepMind/status/2057842131142590512
Biohub releases a world model of protein biology
https://biohub.org/news/world-model-of-protein-biology/
360 drone has arrived. Looking fwd to capturing some big ass gaussian splats.
https://x.com/bilawalsidhu/status/2057635347576524823
Bridging 3D prior from generative models and real 3D scan (only RGB!) is getting sharper everyday!
https://x.com/Almorgand/status/2059011590167392765
Fast 4D Mesh Generation by Spatio-Temporal Attention Chains”” TL;DR: training-free spatio-temporal attention chains generate topology-consistent 4D meshes from video 13× faster while improving temporal correspondence quality
https://x.com/Almorgand/status/2059206856589844552
Large-Scale High-Quality 3D Gaussian Head Reconstruction from Multi-View Captures”” TL;DR: feed-forward UV-parameterized Gaussian head reconstruction scales to 10k+ identities while enabling high-quality avatars and real-time facial animation
https://x.com/Almorgand/status/2059388320078037237
Made a free Pixal3D demo (Tencent’s new image-to-3D model) because I like it a lot 🔥 What’s interesting: pixel-aligned generation: every point in the mesh ties back to a specific input pixel, so silhouettes, textures, and tiny details actually survive into the 3D asset instead
https://x.com/victormustar/status/2057752615396557225
RecGen: 3D Multi-Object Scene Reconstruction from Sparse Observations”” TL;DR: generative 3D scene reconstruction framework that recovers geometry, texture, and pose from sparse RGB-D observations while remaining robust to heavy occlusions and object symmetries
https://x.com/Almorgand/status/2059568924841091282
This is pretty cool — basically “creative upscaling” for 3d scans
https://x.com/bilawalsidhu/status/2059029365002744034
This is text-to-CAD right now. A planetary gear assembly in CAD Explorer where users adjust drive parameters and animation speed, showing gears rotating and meshing realistically. Simulating forces on complex mechanisms strains AI spatial reasoning, yet produces impressive
https://x.com/IlirAliu_/status/2057369823676453305
Velox: Learning Representations of 4D Geometry and Appearance”” TL;DR: learns compact dynamic 4D latent tokens that jointly model geometry and appearance for tasks like video-to-4D generation and 3D tracking
https://x.com/Almorgand/status/2057840305588613505
VGGT-Ω”” TL;DR: scales feed-forward 3D reconstruction to dynamic scenes with memory-efficient register attention, enabling large-scale spatial foundation modeling
https://x.com/Almorgand/status/2057505812868641149
Synthesize Realistic 3D Medical Images at Scale to Ship Pre‑Trained Models | NVIDIA Technical Blog
https://developer.nvidia.com/blog/synthesize-realistic-3d-medical-images-at-scale-to-ship-pre-trained-models/?ncid=so-twit-899791
Would you like to join the research effort on JEPA and World Models easily? After a full year of hard work, we’re excited to finally release stable-worldmodel: an open-source, scalable platform built to accelerate JEPA & World Model research! 📄:
https://x.com/lucasmaes_/status/2060022309759389774
Leave a Reply