Metraipolis › archive › dc_786cfa36fc28

Info3R: Information-Adaptive Test-Time Training for 3D Reconstruction — arXiv:2609.21938

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.21938
License (as archived): CC0-1.0
Archived: 2026-10-10T02:29:45.165Z
SHA-256 of the archived text: 36e5fa0313ff4fa64550fa79c63466b3d604a1379aa15a452c74d6f2ccdaca6b
Stable URL of this page: https://metraipolis.com/documento/dc_786cfa36fc28

How to cite

Info3R: Information-Adaptive Test-Time Training for 3D Reconstruction — arXiv:2609.21938 — arXiv (en), https://arxiv.org/abs/2609.21938 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_786cfa36fc28 (Metraipolis archive dc_786cfa36fc28, sha256 36e5fa0313ff4fa64550fa79c63466b3d604a1379aa15a452c74d6f2ccdaca6b)

BibTeX

@misc{dc786cfa36fc28,
  title = {Info3R: Information-Adaptive Test-Time Training for 3D Reconstruction — arXiv:2609.21938},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.21938},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_786cfa36fc28; sha256 36e5fa0313ff4fa64550fa79c63466b3d604a1379aa15a452c74d6f2ccdaca6b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21938. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Transformer-based models have recently achieved strong performance on 3D reconstruction from images, and recent works extend them to process video streams in an online manner for real-world deployment. However, existing methods overlook two key signals when handling long image streams: the importance of each incoming frame and the information saturation of the model's internal state. In this paper, we propose Info3R, a novel information-adaptive test-time training method for the online 3D reconstruction. We introduce an information-aware state update that modulates the state update strength based on the redundancy and informativeness of each incoming frame. To restore the state's plasticity -- its capacity to incorporate new observations -- we propose a dynamic state reset, triggered by the cumulative magnitude of state updates and the model's prediction confidence and accompanied by an anchor-to-world alignment. Our method achieves consistent improvements on camera pose estimation, video depth estimation, and 3D reconstruction, while substantially mitigating the performance degradation in the long sequence evaluation. Notably, on KITTI Odometry, our method achieves on average 1.68x lower ATE than LongStream, demonstrating its robustness on extended outdoor sequences.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.21938 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-10T02:07:36.110Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).