PulseAugur
实时 09:46:38

UniCorrn Transformer unifies 2D and 3D geometric matching across modalities

Researchers have introduced UniCorrn, a novel Transformer-based model designed to unify geometric matching across various visual data types. This model can handle correspondence tasks between 2D images, 2D images and 3D point clouds, and 3D point clouds themselves, using shared weights. UniCorrn employs a dual-stream decoder to maintain separate appearance and positional features, enabling end-to-end learning and flexible query-based estimation across these heterogeneous modalities. The model demonstrates strong performance, outperforming previous state-of-the-art methods by significant margins on 2D-3D and 3D-3D matching benchmarks. AI

影响 Unifies geometric matching across 2D and 3D data, potentially improving performance on various 3D vision tasks.

排序理由 The cluster describes a new academic paper detailing a novel AI model architecture and its performance on specific benchmarks.

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

UniCorrn Transformer unifies 2D and 3D geometric matching across modalities

报道来源 [2]

  1. arXiv cs.CV TIER_1 English(EN) · Prajnan Goswami, Tianye Ding, Feng Liu, Huaizu Jiang ·

    UniCorrn: Unified Correspondence Transformer Across 2D and 3D

    arXiv:2605.04044v1 Announce Type: new Abstract: Visual correspondence across image-to-image (2D-2D), image-to-point cloud (2D-3D), and point cloud-to-point cloud (3D-3D) geometric matching forms the foundation for numerous 3D vision tasks. Despite sharing a similar problem struct…

  2. arXiv cs.CV TIER_1 English(EN) · Huaizu Jiang ·

    UniCorrn: Unified Correspondence Transformer Across 2D and 3D

    Visual correspondence across image-to-image (2D-2D), image-to-point cloud (2D-3D), and point cloud-to-point cloud (3D-3D) geometric matching forms the foundation for numerous 3D vision tasks. Despite sharing a similar problem structure, current methods use task-specific designs w…