Deep learning is currently the dominant approach to image classification and segmentation, but the performances of deep learning methods are remarkably influenced by the quantity and quality of the ground truth (GT) used for training. In this paper, a deep learning method is presented to deal with the semantic segmentation of very high resolution (VHR) remote sensing data in the case of scarce GT. The main idea is to combine a specific type of deep convolutional neural networks (CNNs), namely, fully convolutional networks (FCNs), with probabilistic graphical models (PGMs). Our method takes advantage of the intrinsic multiscale behavior of FCNs to deal with multiscale data representations and to connect them to a hierarchical Markov model (e.g., making use of a quadtree). As a consequence, the spatial information present in the data is better exploited, allowing a reduced sensitivity to GT incompleteness to be obtained. The marginal posterior mode criterion is used for inference in the proposed framework. To assess the capabilities of the proposed method, the experimental validation is conducted with the ISPRS 2D Semantic Labeling Challenge datasets on the cities of Vaihingen and Potsdam, with some modifications to simulate the spatially sparse GTs that are common in real remote sensing applications. The results are quite significant, as the proposed approach exhibits a higher producer accuracy than the standard FCNs considered and especially mitigates the impact of scarce GTs on minority classes and small spatial details.

In this paper, a hierarchical probabilistic graphical model is proposed to tackle joint classification of multiresolution and multisensor remote sensing images of the same scene. This problem is crucial in the study of satellite imagery and jointly involves multiresolution and multisensor image fusion. The proposed framework consists of a hierarchical Markov model with a quadtree structure to model information contained in different spatial scales, a planar Markov model to account for contextual spatial information at each resolution, and decision tree ensembles for pixelwise modeling. This probabilistic graphical model and its topology are especially fit for application to very high resolution (VHR) image data. The theoretical properties of the proposed model are analyzed: the causality of the whole framework is mathematically proved, granting the use of time-efficient inference algorithms such as the marginal posterior mode criterion, which is non-iterative when applied to quadtree structures. This is mostly advantageous for classification methods linked to multiresolution tasks formulated on hierarchical Markov models. Within the proposed framework, two multimodal classification algorithms are developed, that incorporate Markov mesh and spatial Markov chain concepts. The results obtained in the experimental validation conducted with two datasets containing VHR multispectral, panchromatic, and radar satellite images, verify the effectiveness of the proposed framework. The proposed approach is also compared to previous methods that are based on alternate strategies for multimodal fusion.

This paper presents a method for the automatic learning of the potentials of a stochastic model, in particular a conditional random field (CRF), in a non-parametric fashion. The proposed model is based on a neural architecture, in order to leverage the modeling capabilities of deep learning approaches to directly learn semantic and spatial information from the input data. Specifically, the methodology is based on fully convolutional networks and fully connected neural networks. The idea is to access the multiscale information intrinsically extracted in the intermediate layers of a fully convolutional network through the integration of fully connected neural networks at different scales, while favoring the interpretability of the hidden layers as posterior probabilities. The potentials of the CRF are learned through an additional convolutional layer, whose kernel models the local spatial information considered. The loss function is computed as a linear combination of cross-entropy losses, accounting for the multiscale and the spatial information. To evaluate the capabilities of the proposed approach for the semantic segmentation of remote sensing images, the experimental validation was conducted with the ISPRS 2D Semantic Labeling Challenge Vaihingen and Potsdam datasets and with the IEEE GRSS Data Fusion Contest Zeebruges dataset. As the ground truths of these benchmark datasets are spatially exhaustive, they have been modified to approximate the spatially sparse ground truths common in real remote sensing applications. The results are significant, as the proposed approach obtains higher average classification accuracies than recent state-of-the-art techniques considered in this paper. The code is available at https://github.com/Ayana-Inria/CRFNet-RS.

The growing need for interoperability among the different oceanic monitoring systems to deliver services able to answer the requirements of stakeholders and end-users led to the development of a low-cost machine-to-machine communication system able to guarantee data reliability over marine paths. In this framework, an experimental evaluation of the performance of long-range (LoRa) technology in a fully operational marine scenario has been proposed. In-situ tests were carried out exploiting the availability of (i) a passenger vessel and (ii) a research vessel operating in the Ligurian basin (North-Western Mediterranean Sea) both hosting end-nodes, and (iii) gateways positioned on mountains and hills in the inland areas. Packet loss ratio, packet reception rate, received signal strength indicator, signal to noise, and expected signal power ratio were chosen as metrics in line of sight and not the line of sight conditions. The reliability of Long Range Wide Area Network (LoRaWAN) transmission over the sea has been demonstrated up to more than 110 km in a free space scenario and for more than 20 km in a coastal urban environment.

This paper aims at exploring the potentiality of the multimodal fusion of remote sensing imagery with information coming from mobility demand data in the framework of land-use mapping in urban areas. After a discussion on the function of mobility demand data, a probabilistic fusion framework is developed to take advantage of remote sensing and transport data, and their joint use for urban land-use and land-cover applications in urban and surrounding areas. Two different methods are proposed within this framework, the first based on pixelwise probabilistic decision fusion and the second on the combination with a region-based multiscale Markov random field. The experimental validation is conducted on a case study associated with the city of Genoa, Italy.

Hyperspectral (HS) satellite data are of considerable importance for applications such as environmental monitoring and precision agriculture, given the richness of the spectral information they contain. However, hyperspectral data typically exhibit limited spatial resolution and are less readily available than multispectral (MS) data. This study, which aims to simulate data with high spectral and spatial resolution, explores the use of attention-based spectral reconstruction (SR) techniques, specifically MST++, MIRNet, AWAN and Restormer, to derive HS data in the Visible Near InfraRed (VNIR) and Short-Wave InfraRed (SWIR) from MS imagery. High-resolution MS and HS image pairs are generated from AVIRIS-NG aerial data and employed for training procedures, thereby enabling the reconstruction of HS data that closely resembles the original measurements. The results indicate that SR techniques can considerably enhance the utility of existing MS datasets for HS-dependent applications. Such techniques can effectively be employed to synthesize high-resolution HS data from MS inputs, thereby facilitating the potential for developing a comprehensive end-to-end sensor simulator. This is particularly advantageous in the context of simulating data from a mission that has not yet become operational, as exemplified by the PRISMA-2G data, which could be simulated, for example, from Sentinel-2 data.