StyHighNet: Semi-Supervised Learning Height Estimation from a Single Aerial Image via Unified Style Transferring

• Published in: Sensors (Basel, Switzerland) Vol. 21; no. 7; p. 2272
• Main Authors: Gao, Qian, Shen, Xukun
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 24.03.2021; MDPI
• Subjects: Adaptation; Bias; Cities; Computer vision; convolutional neural network; Datasets; Deep learning; Discrimination; Distribution functions; domain adaptation; Domains; height estimation; Mapping; Remote sensing; Semantics; Semi-supervised learning; style transfer; Teaching methods; height estimation; style transfer; semi-supervised learning; domain adaptation; convolutional neural network

Recovering height information from a single aerial image is a key problem in the fields of computer vision and remote sensing. At present, supervised learning methods have achieved impressive results, but, due to domain bias, the trained model cannot be directly applied to a new scene. In this paper...