Preprint
Sep 2026
FFVO: A Feedforward Pose Decoder for Long-Horizon Visual Odometry
Stable and reliable 4D spatial understanding is fundamental for autonomous driving systems. While feedforward reconstruction networks can estimate camera motion and 3D structure in one pass, pose estimation over long videos remains challenged by computational cost, long-context ambiguity, and temporal instability. To a...
Meng-Li Shih, Shih-Yang Su, Yu-Liang Zou et al.
· 0 citations