Newest papers

Stream4D: Streaming 4D Reconstruction with Training-free Motion Decoupling via Causal KV-Cache Attention

Wei Zhang, Ying Feng, Mengtong Wu, Shiqiang Gong, Huaqi Zhang, Qi Zhang · 2026 · arXiv

DOI: 10.48550/arXiv.2610.31320 · open-access PDF

German

4D-Szenenrekonstruktion erfordert robuste Handhabung von dynamischen Objekten, doch bestehende Streaming-Methoden nehmen üblicherweise statische Szenen an, was zu Posedrift und Ghost-Artifakten über lange Sequenzen führt. Wir führen Stream4D ein, ein trainingsfreies, bewegungsfreies Streaming-Rekonstruktions-Framework, das direkt in kausale Transformers (z. B. StreamVGGT, Stream3R) integriert werden kann. Unsere Schlüsselbeobachtung ist, dass kausale Aufmerksamkeit implizit reiche zeitliche Bewegungssignale im KV-Cache codiert. Wir aggregieren diese Hinweise adaptiv aus mehreren Aufmerksamkeitsschichten und über zeitliche Fenster hinweg und fusionieren

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

Abstract

4D scene reconstruction requires robust handling of dynamic objects, yet existing streaming methods commonly assume static scenes, leading to pose drift and ghosting artifacts over long sequences. We introduce Stream4D, a training-free, motion-decoupled streaming reconstruction framework that can be directly integrated into causal Transformer backbones (e.g., StreamVGGT, Stream3R). Our key observation is that causal attention implicitly encodes rich temporal motion cues within the KV cache. We adaptively aggregate these cues from multiple attention layers and across temporal windows, and fuse

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=2610.31320, 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