Metraipolis › archive › dc_2abbd7822914
Trace4D: Grounded Spatiotemporal Reasoning via 4D State Sequences — arXiv:2611.02759
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.02759
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:42:36.665Z
- SHA-256 of the archived text
973a0f9644fb912b32c54aa88f1f4fcbc0a58fa216c3710a44b2f1d41765f4fb- Stable URL of this page
https://metraipolis.com/documento/dc_2abbd7822914
How to cite
Trace4D: Grounded Spatiotemporal Reasoning via 4D State Sequences — arXiv:2611.02759 — arXiv (en), https://arxiv.org/abs/2611.02759 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2abbd7822914 (Metraipolis archive dc_2abbd7822914, sha256 973a0f9644fb912b32c54aa88f1f4fcbc0a58fa216c3710a44b2f1d41765f4fb)
BibTeX
@misc{dc2abbd7822914,
title = {Trace4D: Grounded Spatiotemporal Reasoning via 4D State Sequences — arXiv:2611.02759},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.02759},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2abbd7822914; sha256 973a0f9644fb912b32c54aa88f1f4fcbc0a58fa216c3710a44b2f1d41765f4fb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02759. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Understanding how objects and their spatial relationships evolve in 3D space over time remains challenging for vision-language models (VLMs), particularly under changing viewpoints. Recent approaches incorporate geometric information into reasoning, but explicit supervision for temporally ordered 4D state evolution and its use as intermediate evidence remains limited. We introduce \textbf{Trace4D}, which trains VLMs to predict 4D state sequences and reason over their inter-timestep changes alongside temporal visual observations. Its 4D-grounded chain-of-thought (CoT) explicitly represents 4D states, analyzes temporal visual context, and connects both as evidence for the answer. We construct Trace4D-Train from in-the-wild videos with 600K 4D perception examples and approximately 70K grounded QA examples, and train with joint supervised fine-tuning followed by geometry-aware GRPO. At inference, Trace4D predicts 4D states and reasoning directly from video without reference annotations or an external geometric model. Experiments on Trace4D-Bench and VLM4D show improved QA accuracy over the base VLMs and benefits from reasoning over 4D state changes and temporal visual observations.
Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.02759 — 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 2026-11-03T11:06:37.586Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).