1 paper
Estelle Chigot, Thomas Oberlin, Manon Huguenin +1
Semantic segmentation networks require large amounts of pixel-level annotated data, which are costly to obtain for real-world images. Computer graphics engines can generate synthet…