arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.09375cs.CV
收藏

DINAR:面向单样本人体化身的神经纹理扩散修复

DINAR: Diffusion Inpainting of Neural Textures for One-Shot Human Avatars

Samsung Research(三星研究院);Cinemersive Labs(Cinemersive实验室)

David Svitov, Dmitrii Gudkov, Renat Bashirov, Victor Lempitsky

AI总结:

本文提出 DINAR,利用在神经纹理空间训练的潜在扩散模型结合 SMPL-X,从单张 RGB 图像逼真重建可动画全身化身,并在 SnapshotPeople 基准上达到最优。

摘要:

我们提出了 DINAR,这是一种根据单张 RGB 图像创建逼真、可绑定全身化身的方法。与以往工作类似,我们的方法使用神经纹理并结合 SMPL-X 人体模型,以实现照片级真实的化身质量,同时使其易于动画化且推理速度快。为恢复纹理,我们使用潜在扩散模型,并展示如何在神经纹理空间中训练该模型。扩散模型的使用使我们能够在给定正面视图的情况下,真实地重建人物背部等较大的不可见区域。我们流程中的模型仅使用 2D 图像和视频进行训练。实验中,我们的方法实现了最先进的渲染质量,并对新姿态和视点具有良好的泛化能力。特别是,该方法在 SnapshotPeople 公共基准上提升了最先进水平。

英文摘要:

We present DINAR, an approach for creating realistic rigged fullbody avatars from single RGB images. Similarly to previous works, our method uses neural textures combined with the SMPL-X body model to achieve photo-realistic quality of avatars while keeping them easy to animate and fast to infer. To restore the texture, we use a latent diffusion model and show how such model can be trained in the neural texture space. The use of the diffusion model allows us to realistically reconstruct large unseen regions such as the back of a person given the frontal view. The models in our pipeline are trained using 2D images and videos only. In the experiments, our approach achieves state-of-the-art rendering quality and good generalization to new poses and viewpoints. In particular, the approach improves state-of-the-art on the SnapshotPeople public benchmark.

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。

↑