Skip to content

Event

PGGT: Feed-Forward Camera Localization with Planar Primitive Grounding

1 reports · 1 sources

Hanqiao Ye and other authors propose the PGGT (Primitive-Grounded Geometry Transformer) network, which takes a single query image and a set of planar map primitives as input and directly predicts a 6-DoF camera pose, achieving feed-forward camera localization, and they claim it can generalize to unseen environments. The authors claim that the method sets new localization performance records on both dense surface maps and minimalist layout models; the code and models have been released, and the paper was accepted to NeurIPS 2026.

Report timeline

  1. PGGT-Loc: Feed-Forward Camera Localization with Primitive-Grounded Geometry Transformer

    The authors propose PGGT (Primitive-Grounded Geometry Transformer), a network that takes a single query image and a set of planar map primitives as input and directly predicts a 6-DoF camera pose, enabling feed-forward camera localization. The method sets new state-of-the-art localization performance on both dense surface maps and minimal layout models. Code and models are publicly available, and the paper was accepted to NeurIPS 2026.

    arXiv cs.RO ·

Related topics

Perception and localizationResearch papers

Interest trend

Current
3
Peak
4Nov 4 01:00
Past 24 hours
–

The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.