CoNeRF: Controllable Neural Radiance Fields

Kania, Kacper; Yi, Kwang Moo; Kowalski, M.; Trzciniski, Tomasz; Tagliasacchi, Andrea

doi:10.1109/cvpr52688.2022.01807

Cited by 52 publications

(36 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…„I M Avatar” [ZAC*21] extends NeRFace based on skinning fields [CZB*21] which are used to deform the canonical NeRF volume given novel expression and pose parameters. CoNeRF [KYK*21] presents a method to disentangle attribute/expression combinations leveraging sparse mask annotations in the training images. They rely on a locality assumption, i.e., one attribute affects only a specific region.…”

Section: Applicationsmentioning

confidence: 99%

Advances in Neural Rendering

Tewari

Thies²,

Mildenhall

et al. 2022

Computer Graphics Forum

151

View full text Add to dashboard Cite

show abstract

Section: Applicationsmentioning

confidence: 99%

Advances in Neural Rendering

Tewari

Thies²,

Mildenhall

et al. 2022

Computer Graphics Forum

151

View full text Add to dashboard Cite

show abstract

“…[Li et al 2017a;Tran and Liu 2018]). Some other methods (e.g., [Guo et al 2021;Kania et al 2022;Park et al 2020Park et al , 2021) use learned embeddings to capture dynamic facial actions, which, however, are hard to interpret except for the correspondence with extra information, such as audio or using an attribute regressor. FENeRF incorporates semantic masks into conditional NeRF [Chan et al 2021b;Schwarz et al 2020] to generate editing interfaces.…”

Section: Neural Face Image Editingmentioning

confidence: 99%

NeRFFaceEditing: Disentangled Face Editing in Neural Radiance Fields

Jiang

Chen

Liu

et al. 2022

SIGGRAPH Asia 2022 Conference Papers

View full text Add to dashboard Cite

Figure 1: Our NeRFFaceEditing method allows users to intuitively edit a facial volume to manipulate its geometry and appearance guided by rendered semantic masks. Given an input sample (a), our method disentangles its geometry and appearance, and allows for one-or multi-label editing. We show a range of flexible face editing tasks that can be achieved with our unified framework: (b) changing the appearance according to a given reference sample while retaining the geometry and 3D consistency; (c) changing the appearance for different views with different reference samples while retaining the geometry; (d) editing multiple labels of the semantic mask for a certain view while keeping the appearance and 3D consistency; (e) editing both the geometry and appearance. The inputs used to control the appearance and geometry are highlighted in green and orange boxes, respectively.

show abstract

“…In contrast to previous sections, geometries change when shape and motion are non-rigid and deformable. This section focuses on methods that extends NeRFs to a dynamic domain, such as a human body with movement or a face with expression [93], [111], [124], [128], [129], [130], [131], [132], [133], [134]. D-NeRF [93] (Nov 2020) takes time as an additional input, and splits the learning process into two main stages: one encoding the scene into a canonical representation and another mapping it into the deformed scene at a particular time.…”

Section: From Static To Dynamicmentioning

confidence: 99%

“…HyperN-eRF achieves excellent results in synthesizing views when scenes contain topological changes, such as a person opening the mouth and a banana being peeled. CoNeRF [134] (Dec 2021), built on HyperNeRF, allows sliders to easily edit face images. Slider values are provided to an MLPparameterized per-attribute Hypermap deformation field.…”

Section: From Static To Dynamicmentioning

confidence: 99%

A Survey on 3D-aware Image Synthesis

Xia¹,

Xue²

2022

Preprint

View full text Add to dashboard Cite

Recent years have seen remarkable progress in deep learning powered visual content creation. This includes 3D-aware generative image synthesis, which produces high-fidelity images in a 3D-consistent manner while simultaneously capturing compact surfaces of objects from pure image collections without the need for any 3D supervision, thus bridging the gap between 2D imagery and 3D reality. The 3D-aware generative models have shown that the introduction of 3D information can lead to more controllable image generation. The task of 3D-aware image synthesis has taken the field of computer vision by storm, with hundreds of papers accepted to top-tier journals and conferences in recent year (mainly the past two years), but there lacks a comprehensive survey of this remarkable and swift progress. Our survey aims to introduce new researchers to this topic, provide a useful reference for related works, and stimulate future research directions through our discussion section. Apart from the presented papers, we aim to constantly update the latest relevant papers along with corresponding implementations at https://weihaox.github.io/awesome-3D-aware-synthesis.

show abstract

CoNeRF: Controllable Neural Radiance Fields

Cited by 52 publications

References 24 publications

Advances in Neural Rendering

Advances in Neural Rendering

NeRFFaceEditing: Disentangled Face Editing in Neural Radiance Fields

A Survey on 3D-aware Image Synthesis

Contact Info

Product

Resources

About