Multi-Task Learning of Height and Semantics from Aerial Images

Abstract

Aerial or satellite imagery is a great source for land surface analysis, which might yield land use maps or elevation models. In this investigation, we present a neural network framework for learning semantics and local height together. We show how this joint multi-task learning benefits to each task on the large dataset of the 2018 Data Fusion Contest. Moreover, our framework also yields an uncertainty map which allows assessing the prediction of the model. Code is available at https://github.com/marcelampc/mtl_aerial_images .

arXiv commentPublished IEEE Geoscience and Remote Sensing Letters. Code https://github.com/marcelampc/mtl_aerial_images

References in corpus (6)

• Adam: A Method for Stochastic Optimization
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• Fully Convolutional Networks for Semantic Segmentation
• GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networks
• Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding
• IM2HEIGHT: Height Estimation from Single Monocular Imagery via Fully Residual Convolutional-Deconvolutional Network

Cited by in corpus (8)

• Semi-Supervised Semantic Segmentation in Earth Observation: The MiniFrance Suite, Dataset Analysis and Multi-task Network Study
• Remote Sensing Novel View Synthesis with Implicit Multiplane Representations
• Estimating optical vegetation indices and biophysical variables for temperate forests with Sentinel-1 SAR data using machine learning techniques: A case study for Czechia
• Seabed-Net: A multi-task network for joint bathymetry estimation and seabed classification from remote sensing imagery in shallow waters
• Geometry-Aware Segmentation of Remote Sensing Images via Implicit Height Estimation
• Aerial Height Prediction and Refinement Neural Networks with Semantic and Geometric Guidance
• Mapping Vulnerable Populations with AI
• Focusing on Shadows for Predicting Heightmaps from Single Remotely Sensed RGB Images with Deep Learning