Reconstructing the World in 3D: From Photosynth to NeRFs and IARPA's Next Bet
bilawalsidhu · x · 2026-08-13
This deep dive traces the evolution of 3D reconstruction from random internet photos.
- History: From Photosynth proving 3D reconstruction of smaller monuments in 2008, to "Building Rome in a Day" scaling it to entire cities in 2009.
- Tech Evolution: Covers the leap from early point clouds to modern NeRFs and 3D Gaussian Splatting, alongside key milestones like the Long Tail Problem, MegaDepth, and the Mannequin Challenge.
- Frontiers: Analyzes applications of MegaScenes, Doppelgangers, foundation models, and the breakthrough of MegaDepth-X.
- Future: Explains why IARPA is funding the next generation of this technology and looks ahead to the 4D frontier and the Sensorium.
Related event: Evolution of 3D Reconstruction from Internet Photos(2 posts)→
More from Multimodal
- Flux 3 Klien first impressions: occasional body horror like SD3, otherwise solid — rionix88 · 2026-10-16
- XGRIDS builds world's largest 3DGS town: 23.5 km² and 6.2 billion Gaussian Splats — bilawalsidhu · 2026-10-16
- IMMERSION: AI film compilation links 12 shorts with one continuous camera move — AI_Andrew · 2026-10-16
- A song with vocals, lyrics and 3D head animation created entirely by Opus 5.5 — repligate · 2026-10-16
- Alibaba open-sources CommerceGen-Studio for commercial-grade e-commerce images — alibabagroup · 2026-10-16
- EVA-Cap: Fudan and OPPO align video captioning with atomic audiovisual events via GRPO — Fudan-University · 2026-10-16