Newest papers

Latent Bridge: From Video Latents to 3D Geometry via Feature Alignment

Shuai Zhang, Zhanpeng Ouyang, Zhengqing Chen, Ruohong Mei, Shaoyu Chen, Weiqiang Ren, Hui Sun, Qian Zhang · 2026 · arXiv

DOI: 10.48550/arXiv.2610.28521 · open-access PDF

German

Video-Generation und feed-forward 3D-Rekonstruktion verwenden in der Regel getrennte Repräsentationen, die RGB-Dekodierung und visuelle Re-Codierung erfordern, um Geometrie von generierten Latenten wiederherzustellen. Wir schlagen Latent Bridge vor, die Video-generierende latente Räume mit geometrischen Feature-Spaces verbindet, um 3D-Geometrie direkt von Video-latenten vorherzusagen. Wir entwickeln Repräsentationsspezifische Brücken für Wan2.1 und V-RAE, die Feed-forward-Vorhersage von Kameras, Tiefe und 3D-Punktwolken ohne den mittleren RGB-Pfad ermöglichen. Jede Brücke wird auf RGB-Videos durch Feature-Destillation trainiert, während

Machine translation (Meta M2M100); the English original follows.

Abstract

Video generation and feed-forward 3D reconstruction typically use separate representations, requiring RGB decoding and visual re-encoding to recover geometry from generated latents. We propose Latent Bridge, which connects video generative latent spaces with geometric feature spaces to directly predict 3D geometry from video latents. We develop representation-specific bridges for Wan2.1 and V-RAE, enabling feed-forward prediction of cameras, depth, and 3D point clouds without the intermediate RGB pathway. Each bridge is trained on RGB videos through feature distillation, while the video encode

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28521, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange