ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-2-2020
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2020, 877–884, 2020
https://doi.org/10.5194/isprs-annals-V-2-2020-877-2020
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2020, 877–884, 2020
https://doi.org/10.5194/isprs-annals-V-2-2020-877-2020

  03 Aug 2020

03 Aug 2020

DISIR: DEEP IMAGE SEGMENTATION WITH INTERACTIVE REFINEMENT

G. Lenczner1,2, B. Le Saux1, N. Luminari2, A. Chan-Hon-Tong1, and G. Le Besnerais1 G. Lenczner et al.
  • 1ONERA / DTIS, Université Paris-Saclay, F-91123 Palaiseau, France
  • 2Delair, FR-31400 Toulouse, France

Keywords: Semantic Segmentation, Deep Neural Networks, Interactive, Aerial Images, Optical imagery, Human-in-the-loop

Abstract. This paper presents an interactive approach for multi-class segmentation of aerial images. Precisely, it is based on a deep neural network which exploits both RGB images and annotations. Starting from an initial output based on the image only, our network then interactively refines this segmentation map using a concatenation of the image and user annotations. Importantly, user annotations modify the inputs of the network – not its weights – enabling a fast and smooth process. Through experiments on two public aerial datasets, we show that user annotations are extremely rewarding: each click corrects roughly 5000 pixels. We analyze the impact of different aspects of our framework such as the representation of the annotations, the volume of training data or the network architecture. Code is available at this address.