SwitchSplat: Learning Discrete Object States from Casually Captured Scenes

Key points

• 3D Gaussian Splatting has made high-fidelity 3D reconstruction fast and accessible, yet existing representations assume a static world.1
• This assumption breaks down in casual captures, where a user naturally interacts with the environment -- opening doors, toggling switches, or moving objects -- introducing discrete state changes that no current framework can model.1
• Time-conditioned dynamic methods such as 4DGS conflate when a state was observed with what state the scene is in, limiting them to replaying seen sequences rather than composing arbitrary object configurations.1
• We validate SwitchSplat on a new casual-capture dataset and show that it outperforms static and dynamic baselines in novel-view synthesis quality, while uniquely supporting rendering of state combinations never seen during capture.1

Sources (1)

• [1]SwitchSplat: Learning Discrete Object States from Casually Captured Scenes
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 17, 06:35 PM

  We present SwitchSplat, the first 3DGS representation to address this gap via a base+slot decomposition: a shared static base captures the invariant scene, while each slot independently holds the two binary states (on/off) of one interactable object.

  3D Gaussian Splatting has made high-fidelity 3D reconstruction fast and accessible, yet existing representations assume a static world.

Extractive summary: sentences quoted from the sources.