Metraipolis › archive › dc_fc6032c42804

Spatial Latent Reasoning for Embodied Reference Understanding — arXiv:2610.09418

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09418
License (as archived)
CC0-1.0
Archived
2026-10-08T12:47:14.930Z
SHA-256 of the archived text
2687fcb819823949a460ad5216d6091a1c9dd4e6156790fe71e553cf2b9ab21c
Stable URL of this page
https://metraipolis.com/documento/dc_fc6032c42804

How to cite

Spatial Latent Reasoning for Embodied Reference Understanding — arXiv:2610.09418 — arXiv (en), https://arxiv.org/abs/2610.09418 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fc6032c42804 (Metraipolis archive dc_fc6032c42804, sha256 2687fcb819823949a460ad5216d6091a1c9dd4e6156790fe71e553cf2b9ab21c)
BibTeX
@misc{dcfc6032c42804,
  title = {Spatial Latent Reasoning for Embodied Reference Understanding — arXiv:2610.09418},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09418},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fc6032c42804; sha256 2687fcb819823949a460ad5216d6091a1c9dd4e6156790fe71e553cf2b9ab21c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09418. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Pointing-gesture visual grounding requires connecting hand geometry with the visual identity and extent of a referred object. A central challenge for continuous latent reasoning is how to organize these complementary cues into useful intermediate supervision. We propose Spatial Latent Reasoning (SLR), a framework that structures this supervision around an ordered sequence of geometric and visual states. A spatial ray state is supervised by fingertip position and pointing direction, followed by four states aligned with target-region features. To construct the visual targets, we introduce parity pooling, which applies polyphase grouping to average region tokens on four interleaved spatial supports. All states are generated recurrently during training and inference; auxiliary annotations are required only during training. On EgoPoint-Ground, the framework improves mIoU over same-backbone supervised fine-tuning by 2.8, 17.5, and 21.1 percentage points on Qwen3.5-4B, Qwen2.5-VL-7B, and Qwen3-VL-8B, respectively, with improvements on both hard subsets. On YouRefIt, it achieves 77.6% precision at IoU 0.5, a numerical margin of 5.2 percentage points over the reported state of the art under differing evaluation protocols. Ablations support joint geometric and visual supervision on the standard and similar-object sets, and favor parity over three alternative pooling operators on the standard set. These results support task-structured supervision for continuous pointing grounding. We will release the code and supporting materials.

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/2610.09418 — 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 2026-10-08T12:18:15.357Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).