We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: CoNeRF: Controllable Neural Radiance Fields

Abstract: We extend neural 3D representations to allow for intuitive and interpretable user control beyond novel view rendering (i.e. camera control). We allow the user to annotate which part of the scene one wishes to control with just a small number of mask annotations in the training images. Our key idea is to treat the attributes as latent variables that are regressed by the neural network given the scene encoding. This leads to a few-shot learning framework, where attributes are discovered automatically by the framework, when annotations are not provided. We apply our method to various scenes with different types of controllable attributes (e.g. expression control on human faces, or state control in movement of inanimate objects). Overall, we demonstrate, to the best of our knowledge, for the first time novel view and novel attribute re-rendering of scenes from a single video.
Comments: Project page: this https URL
Subjects: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
Cite as: arXiv:2112.01983 [cs.CV]
  (or arXiv:2112.01983v2 [cs.CV] for this version)

Submission history

From: Kacper Kania [view email]
[v1] Fri, 3 Dec 2021 15:52:11 GMT (18351kb,D)
[v2] Mon, 6 Dec 2021 07:18:22 GMT (18351kb,D)

Link back to: arXiv, form interface, contact.