Metraipolis › archive › dc_ea31381e9d96

MotionInsight: Diagnosing Object Motion Deficiencies in Generated Videos — arXiv:2609.37030

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.37030
License (as archived)
CC0-1.0
Archived
2026-10-07T16:26:44.057Z
SHA-256 of the archived text
99cfebb3014fdb2017dac9e5820bd163c852ef379ccc8c8c230a787a38a80fc5
Stable URL of this page
https://metraipolis.com/documento/dc_ea31381e9d96

How to cite

MotionInsight: Diagnosing Object Motion Deficiencies in Generated Videos — arXiv:2609.37030 — arXiv (en), https://arxiv.org/abs/2609.37030 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ea31381e9d96 (Metraipolis archive dc_ea31381e9d96, sha256 99cfebb3014fdb2017dac9e5820bd163c852ef379ccc8c8c230a787a38a80fc5)
BibTeX
@misc{dcea31381e9d96,
  title = {MotionInsight: Diagnosing Object Motion Deficiencies in Generated Videos — arXiv:2609.37030},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.37030},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ea31381e9d96; sha256 99cfebb3014fdb2017dac9e5820bd163c852ef379ccc8c8c230a787a38a80fc5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37030. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Despite rapid progress in video generation models, they still exhibit obvious motion deficiencies, often manifested as incorrect object motion. However, most existing video quality evaluations focus on aesthetic quality or text-video alignment. To address this gap, we study object-centric motion fidelity assessment, evaluating target objects along object consistency, motion continuity, and physical plausibility. To achieve this, we first introduce VidMotion, a diagnostic dataset of 6,879 videos with designated moving objects and fine-grained annotations including dimension-wise scores and failure causes. We further propose MotionInsight, a diagnostic evaluator that shifts assessment from implicit RGB-frame observation to explicit motion-space diagnosis. By constructing motion-aware representations, MotionInsight makes subtle motion deficiencies more observable. We also introduce motion-specific rewards during GRPO to transform observed motion into a diagnostic assessment. Experiments demonstrate that MotionInsight provides an effective basis for diagnosing object motion deficiencies, producing human-aligned scores along three dimensions and grounded explanations.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.37030 — 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 2026-10-07T15:54:53.929Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).