研究库 论文知识库

内容库 / 论文

arXiv:2609.25247 · 工程化

Geometric and Semantic Coupling for Interaction Understanding in 3D Scenes

Geometric and Semantic Coupling for Interaction Understanding in 3D Scenes

类型：arxiv
标识：2609.25247
链接：https://arxiv.org/abs/2609.25247
主分类：engineering
形态：method
TLDR：Interaction understanding in 3D scenes requires a joint description of movable parts, their motion, and the regions through which they can be operated. We present Segment-Snap, which connects these outputs through the physical relationship between parts and handles. Learned predictors identify broad part surfaces and small handles. A geometric decoder uses planar and upright priors to constrain motion, then selects hinge lines using predicted handle locations, without training a motion regressor. Conversely, a joint part-and-handle predictor supplies additional handle candidates, whose motion
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-09-23-agent-rag-longcontext-candidates.json

相关 · 按语义匹配

笔记 3DZip · 3D VLM 空间感知特征多样性引导 token 压缩 · 轻量精读与批判

文档 Stephen 自测 · R69 · 2026-09-08

笔记 flyP 轻量精读 · UniVR (arXiv:2607.12800)

笔记 Canvas360 v2 · 几何感知预训练 × 全景图 in-context 生成的深度注入疲劳（重写）

笔记 flyP · 2026-07-22 精读与批判 · Robot-Centric Pointmaps for VLA（具身 VLA + 3D 几何输入重构）

仓库 prakash-kharvi/metaverse-screening-record

arXiv

2609.25247

TLDR

Interaction understanding in 3D scenes requires a joint description of movable parts, their motion, and the regions through which they can be operated. We present Segment-Snap, which connects these outputs through the physical relationship between parts and handles. Learned predictors identify broad part surfaces and small handles. A geometric decoder uses planar and upright priors to constrain motion, then selects hinge lines using predicted handle locations, without training a motion regressor. Conversely, a joint part-and-handle predictor supplies additional handle candidates, whose motion

arXiv ↗ 中文全文