Normal Map Estimation in the Wild

Authors

  • Jorge Condor Lacambra Graphics and Imaging Lab - I3A
  • Manuel Lagunas Arto Graphics and Imaging Lab - I3A
  • Johanna Delanoy ORIGAMI, Laboratoire d’Informatique en Image et Systèmes d’information (LIRIS)
  • Belén Masiá Corcoy Graphics and Imaging Lab - I3A
  • Diego Gutiérrez Pérez Graphics and Imaging Lab - I3A

DOI:

https://doi.org/10.26754/jjii3a.20215977

Abstract

We propose a method to estimate normal maps of objects in the wild, from just a single RGB image. Our approach is based on deep learning, and we use synthetic data to train our network. Lastly, we show its applicability by improving the results of image-based appearance editing tasks.

Downloads

Download data is not yet available.

Downloads

Published

2021-11-12

How to Cite

Condor Lacambra, J., Lagunas Arto, M., Delanoy, J., Masiá Corcoy, B. ., & Gutiérrez Pérez, D. (2021). Normal Map Estimation in the Wild. Jornada De Jóvenes Investigadores Del I3A, 9. https://doi.org/10.26754/jjii3a.20215977

Issue

Vol. 9 (2021): Actas de la X Jornada de Jóvenes Investigadores e Investigadoras del I3A - 21 de octubre de 2021

Section

Artículos (Tecnologías de la Información y las Comunicaciones)