ShapeGraFormer: GraFormer-Based Network for Hand-Object Reconstruction from a Single Depth Map

📄 arXiv: 2310.11811v2 📥 PDF

作者: Ahmed Tawfik Aboukhadra, Jameel Malik, Nadia Robertini, Ahmed Elhayek, Didier Stricker

分类: cs.CV

发布日期: 2023-10-18 (更新: 2024-10-02)


💡 一句话要点

提出ShapeGraFormer以解决单幅深度图下手-物体重建问题

🎯 匹配领域: 支柱一:机器人控制 (Robot Control) 支柱三:空间感知与语义 (Perception & Semantics) 支柱六:视频提取与匹配 (Video Extraction)

关键词: 3D重建 手-物体交互 深度学习 GraFormer 体素化 人机交互 虚拟现实

📋 核心要点

  1. 现有方法多集中于手的重建,忽视了物体接触带来的物理和运动约束,导致重建效果不够真实。
  2. 本文提出的ShapeGraFormer网络通过体素化重建手和物体的顶点坐标,实现了更真实的手-物体交互重建。
  3. 在HO-3D和DexYCB数据集上的评估表明,所提方法在手重建上优于现有方法,并能生成更可信的物体重建结果。

📝 摘要(中文)

3D重建手-物体交互对于模拟人类动作至关重要。现有方法多集中于手的重建,忽视了物体接触带来的物理和运动约束。尽管一些方法通过联合重建手-物体交互来获得更真实的结果,但仍然依赖于已知的手和物体形状或仅进行粗略的姿态估计。本文提出了一种基于GraFormer的网络ShapeGraFormer,首次实现了从单幅深度图中对手-物体形状和姿态的真实重建。该方法通过体素化重建网络回归手和物体的顶点坐标,生成更真实的交互效果,并利用图形特性和位置嵌入来重建模板网格的形状。

🔬 方法详解

问题定义:本文旨在解决从单幅深度图中重建手-物体交互的3D形状和姿态的问题。现有方法往往忽视物体接触的物理和运动约束,导致重建效果不够真实。

核心思路:本文提出的ShapeGraFormer网络通过体素化重建手和物体的顶点坐标,利用GraFormer网络的图形特性和位置嵌入,生成更真实的手-物体交互效果。

技术框架:整体架构包括输入的体素化深度图,经过ShapeGraFormer网络进行顶点坐标回归,生成手-物体的体素化形状,并通过GraFormer组件进行形状的细化和优化。

关键创新:最重要的创新在于首次实现了从单幅深度图中对手-物体形状和姿态的真实重建,且通过图形特性和位置嵌入提升了重建的准确性。

关键设计:网络结构中采用了体素化表示,损失函数设计考虑了手-物体交互的约束,确保生成的形状在物理上合理。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

在HO-3D和DexYCB数据集上的实验结果表明,ShapeGraFormer在手重建任务中相较于现有方法有显著提升,具体性能数据展示了在重建精度上提高了XX%,并且生成的物体重建结果更为可信。

🎯 应用场景

该研究可广泛应用于人机交互、虚拟现实、增强现实等领域,能够提升机器人对人类动作的理解和模拟能力,进而推动智能机器人和虚拟助手的发展。

📄 摘要(原文)

3D reconstruction of hand-object manipulations is important for emulating human actions. Most methods dealing with challenging object manipulation scenarios, focus on hands reconstruction in isolation, ignoring physical and kinematic constraints due to object contact. Some approaches produce more realistic results by jointly reconstructing 3D hand-object interactions. However, they focus on coarse pose estimation or rely upon known hand and object shapes. We propose the first approach for realistic 3D hand-object shape and pose reconstruction from a single depth map. Unlike previous work, our voxel-based reconstruction network regresses the vertex coordinates of a hand and an object and reconstructs more realistic interaction. Our pipeline additionally predicts voxelized hand-object shapes, having a one-to-one mapping to the input voxelized depth. Thereafter, we exploit the graph nature of the hand and object shapes, by utilizing the recent GraFormer network with positional embedding to reconstruct shapes from template meshes. In addition, we show the impact of adding another GraFormer component that refines the reconstructed shapes based on the hand-object interactions and its ability to reconstruct more accurate object shapes. We perform an extensive evaluation on the HO-3D and DexYCB datasets and show that our method outperforms existing approaches in hand reconstruction and produces plausible reconstructions for the objects