AI RESEARCH
TALO: Pushing 3D Vision Foundation Models Towards Globally Consistent Online Reconstruction
arXiv CS.CV
•
ArXi:2512.02341v3 Announce Type: replace 3D vision foundation models have shown strong generalization in reconstructing key 3D attributes from uncalibrated images through a single feed-forward pass. However, when deployed in online settings such as driving scenarios, predictions are made over temporal windows, making it non-trivial to maintain consistency across time. Recent strategies align consecutive predictions by solving global transformation, yet our analysis reveals their fundamental limitations in assumption validity, local alignment scope, and robustness under noisy geometry.