arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2403.12011cs.CV

HOIDiffusion:生成逼真的3D手-物交互数据

HOIDiffusion: Generating Realistic 3D Hand-Object Interaction Data

UC San Diego(加州大学圣迭戈分校);NVIDIA(英伟达公司)

Mengqi Zhang, Yang Fu, Zheng Ding, Sifei Liu, Zhuowen Tu, Xiaolong Wang

AI总结:

针对3D手-物交互数据稀缺问题,提出条件扩散模型HOIDiffusion,以几何结构和文本为输入生成逼真数据,并用于提升6D姿态估计。

摘要:

由于在扩展数据收集过程中的硬件限制,3D手-物交互数据十分稀缺。本文提出HOIDiffusion用于生成逼真且多样的3D手-物交互数据。我们的模型是一个条件扩散模型,它以3D手-物几何结构和文本描述作为输入进行图像合成。这提供了更可控且逼真的合成,因为我们可以以解耦的方式指定结构和风格输入。HOIDiffusion通过利用在大规模自然图像上预训练的扩散模型和少量3D人体演示进行训练。除了可控图像合成外,我们采用生成的3D数据来学习6D物体姿态估计,并展示了其在改进感知系统方面的有效性。项目页面:https://mq-zhang1.github.io/HOIDiffusion

英文摘要:

3D hand-object interaction data is scarce due to the hardware constraints in scaling up the data collection process. In this paper, we propose HOIDiffusion for generating realistic and diverse 3D hand-object interaction data. Our model is a conditional diffusion model that takes both the 3D hand-object geometric structure and text description as inputs for image synthesis. This offers a more controllable and realistic synthesis as we can specify the structure and style inputs in a disentangled manner. HOIDiffusion is trained by leveraging a diffusion model pre-trained on large-scale natural images and a few 3D human demonstrations. Beyond controllable image synthesis, we adopt the generated 3D data for learning 6D object pose estimation and show its effectiveness in improving perception systems. Project page: https://mq-zhang1.github.io/HOIDiffusion

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑