Event
After feedforward reconstruction becomes 1200× faster, dynamic 4D reconstruction still needs BA
1 reports · 1 sources
StreamSplat at ICLR 2026 achieves purely feedforward, online generation of dynamic 3DGS directly from uncalibrated video streams, at 1200× the speed of optimization-based methods. However, MoRe at CVPR 2026 actively embeds a BA-like refinement module in its purely feedforward pipeline; after removing this module, Sintel's ATE rises from 0.147 to 0.155, indicating that BA-style refinement can still improve accuracy. BA-Track, an ICCV 2025 Oral, uses a 3D point tracker to decompose camera and object motion, adapting classical BA to dynamic scenes. These works show that feedforward speedups have not replaced BA; rather, they have shifted BA's role from the main solver to a component for refinement or adaptation to dynamic scenes.
Report timeline
- After feedforward reconstruction becomes 1,200 times faster, is BA still necessary in dynamic 4D reconstruction?
StreamSplat at ICLR 2026 achieves purely feedforward, online generation of dynamic 3DGS directly from uncalibrated video streams, at 1,200 times the speed of optimization-based methods. But MoRe at CVPR 2026 actively embeds a BA-like refinement module in its purely feedforward pipeline; removing it raises Sintel ATE from 0.147 to 0.155. BA-Track, an ICCV 2025 Oral, uses a 3D point tracker to decompose camera and object motion, adapting classic BA to dynamic scenes.
3D视觉工坊 · 微信公众号 ·