1 Hit in 0.69 sec

NeSF: Neural Semantic Fields for Generalizable Semantic Segmentation of 3D Scenes [article]

Suhani Vora and Noha Radwan and Klaus Greff and Henning Meyer and Kyle Genova and Mehdi S. M. Sajjadi and Etienne Pot and Andrea Tagliasacchi and Daniel Duckworth
2021 arXiv   pre-print
We present NeSF, a method for producing 3D semantic fields from posed RGB images alone.  ...  Despite being trained on 2D signals alone, our method is able to generate 3D-consistent semantic maps from novel camera poses and can be queried at arbitrary 3D points.  ...  Klaus Greff was responsible for Kubric [45], the technology used to generate the datasets in this work.  ... 
arXiv:2111.13260v3 fatcat:r3kbtzi545cr5olbixmpyux2yi