Showing cs.CVShow all
2 papers · 1 filter
cs.CV2026
F SLAM: Turning Feed-Forward Geometry into Persistent Factors for SLAM
Zhisong Xu, Fan Zhu, Jiawei Qian +3
Feed-forward 3D models provide strong multi-view geometric priors, while on- line simultaneous localization and mapping (SLAM) relies mainly on local mea- surements and can accumul…
cs.CV2026
FrameVGGT: Coherence-Preserving Memory for Bounded Streaming Geometry
Zhisong Xu, Takeshi Oishi
Streaming Visual Geometry Transformers such as StreamVGGT enable strong online 3D perception, but their KV-cache grows unbounded over long streams, limiting practical deployment. W…