CVPR 2026 Best Paper: D4RT enables ultra-fast 4D scene reconstruction
andrew_n_carr · x · 2026-08-30
D4RT, by Google DeepMind et al., won the CVPR 2026 Best Paper. The model uses a unified transformer architecture to jointly infer depth, spatio-temporal correspondence, and full camera parameters from video.
Its core innovation is a novel querying mechanism that sidesteps heavy dense per-frame decoding. The decoding interface allows flexible probing of any 3D point in space and time. This results in highly efficient training and inference, achieving 200+ FPS pose estimation and setting a new state of the art across 4D reconstruction tasks.
More from Multimodal
- Halcyon: no text pretraining, beats LlamaGen-XL/8 with 18× less compute — JosephJacks_ · 2026-10-16
- Halcyon early-fusion image models match late-fusion with 275x less pretraining compute, open-sourced — davidwromero · 2026-10-16
- Running Wan2.1 14B FP16 on a 12GB laptop GPU: VRAM swap stress test finishes in 4m11s — RYuu-1120 · 2026-10-16
- Striking Midjourney generations: black holes and planetary aesthetics — ciguleva · 2026-10-16
- Fable 5.5 may drop today as Andrew Curran hints at early release — kimmonismus · 2026-10-16
- Kling 4.0 Adds Timeline Editing, Endorsed by 'Seoul Spring' Director Kim Seong-su — xiaohu · 2026-10-16