Metraipolis › archive › dc_5da40e51253a

ProMo3D: Probing Motion Cues in Frozen 3D Foundation Models — arXiv:2610.33802

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.33802
License (as archived)
CC0-1.0
Archived
2026-10-27T08:00:30.715Z
SHA-256 of the archived text
8cc63c05b206f6ae02a680ccc1c2ab2f1212d6f22ad57163179a2af23f68ab78
Stable URL of this page
https://metraipolis.com/documento/dc_5da40e51253a

How to cite

ProMo3D: Probing Motion Cues in Frozen 3D Foundation Models — arXiv:2610.33802 — arXiv (en), https://arxiv.org/abs/2610.33802 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5da40e51253a (Metraipolis archive dc_5da40e51253a, sha256 8cc63c05b206f6ae02a680ccc1c2ab2f1212d6f22ad57163179a2af23f68ab78)
BibTeX
@misc{dc5da40e51253a,
  title = {ProMo3D: Probing Motion Cues in Frozen 3D Foundation Models — arXiv:2610.33802},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.33802},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5da40e51253a; sha256 8cc63c05b206f6ae02a680ccc1c2ab2f1212d6f22ad57163179a2af23f68ab78},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.33802. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Feed-forward 3D foundation models have recently reshaped multi-view reconstruction, yet dynamic 4D understanding still often relies on large-scale motion supervision or hand-crafted, architecture-specific heuristics. We ask whether frozen 3D foundation model tokens already encode motion-relevant information that can be exposed by lightweight readout functions, rather than re-learning motion-specific features with high-capacity decoders. Through probing on three representative backbones, we find that static--dynamic separation is decodable from frozen token representations. Our rank-restricted and spectral analyses on $\pi^3$X reveal that the recovered motion cues are concentrated in a low-dimensional subspace. Building on this, we propose Probing-based Motion Distillation (PMD), which converts pairwise motion-state compatibility scores against a fixed anchor token into dense motion masks. Trained on only 160 YouTube-VOS videos using a single 16 GB GPU, PMD transfers zero-shot to DAVIS, SegTrackv2, and FBMS-59 without target-benchmark fine-tuning. With multi-view pretrained geometric backbones, PMD outperforms training-free heuristics and approaches substantially heavier supervised and self-supervised methods, suggesting that motion decoding from frozen 3D foundation models can be lightweight and data-efficient.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.33802 — 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 2026-10-27T07:55:30.855Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).