Language-driven Object Fusion into Neural Radiance Fields with Pose-Conditioned Dataset Updates

arXiv:2309.11281v3 Announce Type: replace
Abstract: Neural radiance field is an emerging rendering method that generates high-quality multi-view consistent images from a neural scene representation and volume rendering. Although neural radiance field-based techniques are robust for scene reconstruction, their ability to add or remove objects remains limited. This paper proposes a new language-driven approach for object manipulation with neural radiance fields through dataset updates. Specifically, to insert a new foreground object represented by a set of multi-view images into a background …

abstract arxiv consistent cs.cv dataset fields fusion images language neural radiance field neural radiance fields neural scene representation object objects paper quality rendering representation robust type updates view