Salvato in:
Dettagli Bibliografici
Autori principali: Deng, Yu, Wang, Duomin, Ren, Xiaohang, Chen, Xingyu, Wang, Baoyuan
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:https://arxiv.org/abs/2311.18729
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909215025528832
author Deng, Yu
Wang, Duomin
Ren, Xiaohang
Chen, Xingyu
Wang, Baoyuan
author_facet Deng, Yu
Wang, Duomin
Ren, Xiaohang
Chen, Xingyu
Wang, Baoyuan
contents Existing one-shot 4D head synthesis methods usually learn from monocular videos with the aid of 3DMM reconstruction, yet the latter is evenly challenging which restricts them from reasonable 4D head synthesis. We present a method to learn one-shot 4D head synthesis via large-scale synthetic data. The key is to first learn a part-wise 4D generative model from monocular images via adversarial learning, to synthesize multi-view images of diverse identities and full motions as training data; then leverage a transformer-based animatable triplane reconstructor to learn 4D head reconstruction using the synthetic data. A novel learning strategy is enforced to enhance the generalizability to real images by disentangling the learning process of 3D reconstruction and reenactment. Experiments demonstrate our superiority over the prior art.
format Preprint
id arxiv_https___arxiv_org_abs_2311_18729
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Portrait4D: Learning One-Shot 4D Head Avatar Synthesis using Synthetic Data
Deng, Yu
Wang, Duomin
Ren, Xiaohang
Chen, Xingyu
Wang, Baoyuan
Computer Vision and Pattern Recognition
Existing one-shot 4D head synthesis methods usually learn from monocular videos with the aid of 3DMM reconstruction, yet the latter is evenly challenging which restricts them from reasonable 4D head synthesis. We present a method to learn one-shot 4D head synthesis via large-scale synthetic data. The key is to first learn a part-wise 4D generative model from monocular images via adversarial learning, to synthesize multi-view images of diverse identities and full motions as training data; then leverage a transformer-based animatable triplane reconstructor to learn 4D head reconstruction using the synthetic data. A novel learning strategy is enforced to enhance the generalizability to real images by disentangling the learning process of 3D reconstruction and reenactment. Experiments demonstrate our superiority over the prior art.
title Portrait4D: Learning One-Shot 4D Head Avatar Synthesis using Synthetic Data
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.18729