Search

The quantification of emissions of greenhouse gases and air pollutants through the inversion of plumes in satellite images remains a complex problem that current methods can only assess with significant uncertainties. The anticipated launch of the CO2M (Copernicus Anthropogenic Carbon Dioxide Monitoring) satellite constellation in 2026 is expected to provide high-resolution images of CO2 (carbon dioxide) column-averaged mole fractions (XCO2), opening up new possibilities. However, the inversion of future CO2 plumes from CO2M will encounter various obstacles. A challenge is the low CO2 plume signal-to-noise ratio due to the variability in the background and instrumental errors in satellite measurements. Moreover, uncertainties in the transport and dispersion processes further complicate the inversion task. To address these challenges, deep learning techniques, such as neural networks, offer promising solutions for retrieving emissions from plumes in XCO2 images. Deep learning models can be trained to identify emissions from plume dynamics simulated using a transport model. It then becomes possible to extract relevant information from new plumes and predict their emissions. In this paper, we develop a strategy employing convolutional neural networks (CNNs) to estimate the emission fluxes from a plume in a pseudo- XCO2 image. Our dataset used to train and test such methods includes pseudo-images based on simulations of hourly XCO2 , NO2 (nitrogen dioxide), and wind fields near various power plants in eastern Germany, tracing plumes from anthropogenic and biogenic sources. CNN models are trained to predict emissions from three power plants that exhibit diverse characteristics. The power plants used to assess the deep learning model's performance are not used to train the model. We find that the CNN model outperforms state-of-the-art plume inversion approaches, achieving highly accurate results with an absolute error about half of that of the cross-sectional flux method and an absolute relative error of ∼ 20 % when only the XCO2 and wind fields are used as inputs. Furthermore, we show that our estimations are only slightly affected by the absence of NO2 fields or a detection mechanism as additional information. Finally, interpretability techniques applied to our models confirm that the CNN automatically learns to identify the XCO2 plume and to assess emissions from the plume concentrations. These promising results suggest a high potential of CNNs in estimating local CO2 emissions from satellite images. [ABSTRACT FROM AUTHOR]

The quantification of emissions of greenhouse gases and air pollutants through the inversion of plumes in satellite images remains a complex problem that current methods can only assess with significant uncertainties. The anticipated launch of the CO2M satellite constellation in 2026 is expected to provide high-resolution images of CO2 column-averaged mole fractions (XCO2), opening up new possibilities. However, the inversion of future CO2 plumes from CO2M will encounter various obstacles. A challenge is the CO2 plume low signal-to-noise ratio, due to the variability of the background and instrumental errors in satellite measurements. Moreover, uncertainties in the transport and dispersion processes further complicate the inversion task. To address these challenges, deep learning techniques, such as neural networks, offer promising solutions for retrieving emissions from plumes in XCO2 images. Deep learning models can be trained to identify emissions from plume dynamics simulated using a transport model. It then becomes possible to extract relevant information from new plumes and predict their emissions. In this paper, we employ convolutional neural networks (CNN) to estimate the emission fluxes from a plume in a pseudo XCO2 image. Our dataset used to train and test such methods includes pseudo images based on simulations of hourly XCO2, NO2 and wind fields near various power plants in Eastern Germany, tracing plumes from anthropogenic and biogenic sources. CNN models are trained to predict emissions from three power plants that exhibit diverse characteristics. The power plants used to assess the deep learning model's performance are not used to train the model. We find that the CNN model outperforms state of the art plume inversion approaches, achieving highly accurate results with an absolute error about half of that of the cross-sectional flux method. Furthermore, we show that our estimations are only slightly affected by the absence of NO2 fields or a detection mechanism as additional information. Finally, interpretability techniques applied to our models confirm that the CNN automatically learns to identify the XCO2 plume and to assess emissions from the plume concentrations. These promising results suggest a high potential of CNNs in estimating local CO2 emissions from satellite images. [ABSTRACT FROM AUTHOR]

The utilization of stationary underwater cameras is a modern and well-adapted approach to provide a continuous and cost-effective long-term solution to monitor underwater habitats of particular interest. A common goal of such monitoring systems is to gain better insight into the dynamics and condition of populations of various marine organisms, such as migratory or commercially relevant fish taxa. This paper describes a complete processing pipeline to automatically determine the abundance, type and estimate the size of biological taxa from stereoscopic video data captured by the stereo camera of a stationary Underwater Fish Observatory (UFO). A calibration of the recording system was carried out in situ and, afterward, validated using the synchronously recorded sonar data. The video data were recorded continuously for nearly one year in the Kiel Fjord, an inlet of the Baltic Sea in northern Germany. It shows underwater organisms in their natural behavior, as passive low-light cameras were used instead of active lighting to dampen attraction effects and allow for the least invasive recording possible. The recorded raw data are pre-filtered by an adaptive background estimation to extract sequences with activity, which are then processed by a deep detection network, i.e., Yolov5. This provides the location and type of organisms detected in each video frame of both cameras, which are used to calculate stereo correspondences following a basic matching scheme. In a subsequent step, the size and distance of the depicted organisms are approximated using the corner coordinates of the matched bounding boxes. The Yolov5 model employed in this study was trained on a novel dataset comprising 73,144 images and 92,899 bounding box annotations for 10 categories of marine animals. The model achieved a mean detection accuracy of 92.4%, a mean average precision (mAP) of 94.8% and an F1 score of 93%. [ABSTRACT FROM AUTHOR]

During recent years, maintaining scientific and medical databases has turned out to be a major challenge in the medical sector. The medical data regarding the patients comprise diverse diagnostic medical reports and features for such disease need to be recorded carefully for providing high-quality treatments. Therefore, this paper proposes a hybrid deep convolutional neural network (HDCNN)-based COOT approach to predict the risk of diseases. Initially, an improved crossover-based levy flight optimization algorithm (ICLFDO) algorithm is employed in selecting and processing the unstructured textual data. Then later, the HDCNN-COOT approach predicts the diseases more accurately. In addition to this, the classifier predicts whether the patient is at risk of disease in the future or not. The efficiency of the proposed model is evaluated by using a dataset obtained from University Hospital of Ludwig Maximilian University of Munich, Germany. The data collected in the data center include 29,477,035 data items from 36,082 patients. The proposed model is capable of using structured data and uses textual data of the patients. Finally, the higher classification accuracy and improved performance of the classifiers can be viewed using the experimental results conducted on five different datasets. [ABSTRACT FROM AUTHOR]

Background This paper explores the role of providing structure for students within a new blended-learning family medicine curriculum at Saarland University. This study intends to clarify the role of different structural elements for different learning approaches (deep vs. structured vs. surface learning). Methods This study is based on a mixed-methods approach. Fifth year medical students' (n = 84) learning approaches, measured by a shortened version of the ASSIST questionnaire, were correlated with students' academic performance in state exams and the family medicine exam. Based on learning approaches and age, participants were purposefully sampled into two semi-structured group interviews with seven participants each. Results Although surface approach learning weakly correlated with lower performance in the first state exam (rho = 0.231), surface approach learning did not correlate with lower performance in the family medicine assessment. Based on qualitative results, the structural design of the curriculum seemed to have a high impact on students' satisfaction and motivation to learn. The reduction of extraneous cognitive load was identified as effective in motivating students to study. Deep-learning students highlighted the positive effects of curriculum activities that foster germane cognitive load through schema- based learning. Conclusions Especially for surface approach learners it may be important to provide structured guidance and provide a clear alignment of course content in a blended-learning curriculum, even in higher education. Deep and structured approach learners also benefit from a clear course structure that keeps extraneous cognitive load low and fosters schema acquisition. [ABSTRACT FROM AUTHOR]

During the lockdown of universities and the COVID-Pandemic most students were restricted to their homes. Novel and instigating teaching methods were required to improve the learning experience and so recent implementations of the annual PhysioNet/Computing in Cardiology (CinC) Challenges posed as a reference. For over 20 years, the challenges have proven repeatedly to be of immense educational value, besides leading to technological advances for specific problems. In this paper, we report results from the class ‘Artificial Intelligence in Medicine Challenge’, which was implemented as an online project seminar at Technical University Darmstadt, Germany, and which was heavily inspired by the PhysioNet/CinC Challenge 2017 ‘AF Classification from a Short Single Lead ECG Recording’. Atrial fibrillation is a common cardiac disease and often remains undetected. Therefore, we selected the two most promising models of the course and give an insight into the Transformer-based DualNet architecture as well as into the CNN-LSTM-based model and finally a detailed analysis for both. In particular, we show the model performance results of our internal scoring process for all submitted models and the near state-of-the-art model performance for the two named models on the official 2017 challenge test set. Several teams were able to achieve F1 scores above/close to 90% on a hidden test-set of Holter recordings. We highlight themes commonly observed among participants, and report the results from the self-assessed student evaluation. Finally, the self-assessment of the students reported a notable increase in machine learning knowledge. [ABSTRACT FROM AUTHOR]

• Soil reflectance composites were generated with Sentinel-2 L2A data. • A multi-channel CNN learning framework for soil organic carbon prediction is proposed. • A dual input deep learning architecture improves the predictions. • The framework produced results that can be interpreted. The Copernicus Sentinel-2 multispectral imagery data may be aggregated to extract large-scale, bare soil, reflectance composites, which enable soil mapping applications. In this paper, this approach was tested in the German federal state of Bavaria, to provide estimations for soil organic carbon (SOC). Different temporal ranges were considered for the generation of the composites, including multi-annual and seasonal ranges. A novel multi-channel convolutional neural network (CNN) is proposed. By leveraging the advantages of deep learning techniques, it utilizes complementary information from different spectral pre-treatment techniques. The SOC predictions indicated little dissimilarity amongst the different composites, with the best performance attained for the six-year composite containing only spring months (RMSE = 12.03 g C · kg−1, R2 = 0.64, RPIQ = 0.89). It has been demonstrated that these outcomes outperform other well-known machine learning techniques. An ablation analysis was accordingly performed to evaluate the interplay of the CNN's different components to disentangle the advantages of each aspect of the proposed framework. Finally, a DUal inPut deep LearnIng architecture, named DUPLICITE, is proposed, which concatenates deep spectral features derived from the CNN mentioned earlier, as well as topographical and environmental covariates through an artificial neural network (ANN) to exploit their complementarity. The proposed approach was demonstrated to provide an improvement in the overall prediction performance (RMSE = 11.60 gC · kg−1, R2 = 0.67, RPIQ = 0.92). [ABSTRACT FROM AUTHOR]

The increase of electric vehicles (EVs) in the power system, in addition to environmental benefits, also has economic advantages. Charging parking lots can perform better economically through aggregators the benefits of participating in energy markets and developing different charging programs. The main challenge for participating in the electricity market for parking lots is the uncertainties in the behavior of EV owners and the day-ahead market (DAM) price. Therefore, this paper proposes a deep learning-based framework for day-ahead optimal charging scheduling, which is in three stages. In the first stage, the parameters of charging sessions are modeled using the proposed Copula generative adversarial network (CopulaGAN). In the second stage, the DAM price is forecasted using the time series multi-input nonlinear autoregressive neural network for 24 h ahead. In addition, in the proposed framework, the balancing energy market (BEM) is used to keep the charging power balance on the implementation day of the optimal charging program. In the third stage, optimal day-ahead EV scheduling is performed. The simulation is carried out using Grey Wolf Optimization (GWO) based on the Caltech charging sessions dataset and the Germany energy market dataset. The simulation results show that using the proposed deep learning framework reduces the charging cost for the parking lots compared to the fixed price approaches, the use of the BEM, the time of use (TOU) program, and using normal distribution modeling about 19.94 €, 159.97 €, 14.27 €, and 114.03 € respectively. • Presenting a deep learning framework for modeling day-ahead scheduling. • The proposed framework is applied to Caltech and the German energy market dataset. • Results show a 15% reduction in charging costs compared to conventional methods. [ABSTRACT FROM AUTHOR]

The estimation of the electrical power output and energy yield of high concentration photovoltaic (HCPV) modules is a hard task because of the many parameters involved. In this work, we propose the usage of an Artificial Neural Networks (ANN) method to estimate the maximum power output of a FLATCON® CPV-module manufactured by Fraunhofer ISE. The advantage of the ANN is that it is a part of the Artificial Intelligence Deep Learning domain, which makes it easy to model complex systems. A detailed knowledge about the underlying module technology is not necessary. In this work, various meteorological parameters are tested as inputs of the ANN including spectral matching ratios for considering the spectral variation of the solar irradiance to evaluate the maximum power output of the HCPV module under study. Eleven scenarios considering different combinations of input parameters have been investigated. It is demonstrated that the ANN gives excellent results and allows for an accurate prediction of the HCPV module's instantaneous power output with only a few amount of data used on training the ANN. The model has been tested using the measured data set of a FLATCON® CPV-module located on the rooftop of the Fraunhofer ISE in Freiburg, Germany. The module has been electrically characterized outdoors in the period from April 2013 to April 2014. The accuracy of the proposed ANN approach was analyzed using this measurement data in combination with error metrics. It was found that the best agreement between the measured and the predicted maximum power output was achieved when using direct normal irradiance, wind speed, ambient temperature and spectral matching ratios as input. The normalized root mean square error for the maximum power output over one year is found to be between 2.2 and 4.6%. The deviation of the modelled energy yield to the measured one is in the range of 0.2–2.2%. • We use ANN method to predict the maximum power output of a FLATCON® CPV-module. • The accuracy of model used was validated through measurements and metrics errors. • The spectral effect of DNI through the use of SMRs has improved the model performance. • The knowledge of module technology is not necessary for performance prediction. • Only small amount of data are needed for performance prediction using this approach. Short Abstract: This paper uses the artificial Neural Networks model to estimate the electrical power and the energy yield of high concentration photovoltaic modules. The model has been verified using a whole year measurement data of power output of FLATCON CPV Module manufactured by Fraunhofer ISE, Germany. A best agreement between the measured and computed energy output was found trough the analysis of eleven scenarios where different input parameters were combined. [ABSTRACT FROM AUTHOR]

Solar power integration has shown a significant growth in many power systems during the last decade. The intermittent nature of solar irradiance tends to vary the amount of solar power in the system and an accurate solar power forecasting method can be used to tackle this in power system planning and operation. In this paper, authors have proposed a generalized ensemble model integrating deep learning techniques to generate accurate solar power forecasts for 21 solar photovoltaic facilities located in Germany. Most important weather parameters for solar power generation are selected through a feature selection process. In addition, a weather classification approach is used to cluster the dataset and for each cluster, a separate ensemble algorithm is assigned. Finally, considering the prediction errors in each cluster, a novel ensemble model is developed. The proposed models are evaluated using root mean square error and results are compared with single machine learning techniques and available forecasting models in the literature. Compared to deep belief network, support vector regression and random forest regression models, the proposed ensemble model with cloud classification reduces RMSE error by 10.49%, 7.78%, and 7.95% respectively. Results show that the weather classification approach reduces the forecasting error by a considerable margin and the proposed ensemble model provides a better forecasting accuracy than single machine learning methods. [ABSTRACT FROM AUTHOR]

Today's methods for data acquisition usually involve a considerable amount of manual work and, depending on the level of detail, must find a compromise between accuracy and workload. As a result, various typologies of building stocks have been developed. These typologies provide engineers and architects with rough descriptions for classifying buildings into certain periods and construction types. However, real-world applications lack objective decision-making, as these classifications are usually conducted by manual investigations of buildings. To overcome this subjective approach and to ensure the reproducibility of data acquisition, this paper proposes an image-based classification using Deep Learning (DL) methods. We present the acquisition of ground truth and describe the training and testing routines on a dataset of 5,307 images obtained in the city of Weimar (located in the Free State of Thuringia, Germany), a mid-sized city representative of eastern Germany. All building images are linked to an existing building typology, allowing a classification of buildings in ten predefined and acknowledged periods. These periods are ranging from pre-1859 to post-2002. As a benchmark, an expert human baseline is acquired, showing a lower prediction accuracy than the results obtained by DL methods. Our results highlight the potential for automation in image analysis and beneficial impacts for Building Energy Modeling (BEM) on urban scale. Thus, a trustworthy data acquisition regarding the building age holds benefits for various stakeholders in BEM (engineers, architects, policymakers as well as customers). • Two CNN-architectures (ResNet50 and AlexNet) are used for image-classification. • Extraction of annotated image-patches facilitates an enhancement of the image set. • Quantity of training images and applied optimizer significantly influence results. • CNN-based estimations successfully linked to existing and popular building typology. • Accuracies of CNN-based estimation of building age outperform the human baseline. [ABSTRACT FROM AUTHOR]

Abstract Background Several recent publications have demonstrated the use of convolutional neural networks to classify images of melanoma at par with board-certified dermatologists. However, the non-availability of a public human benchmark restricts the comparability of the performance of these algorithms and thereby the technical progress in this field. Methods An electronic questionnaire was sent to dermatologists at 12 German university hospitals. Each questionnaire comprised 100 dermoscopic and 100 clinical images (80 nevi images and 20 biopsy-verified melanoma images, each), all open-source. The questionnaire recorded factors such as the years of experience in dermatology, performed skin checks, age, sex and the rank within the university hospital or the status as resident physician. For each image, the dermatologists were asked to provide a management decision (treat/biopsy lesion or reassure the patient). Main outcome measures were sensitivity, specificity and the receiver operating characteristics (ROC). Results Total 157 dermatologists assessed all 100 dermoscopic images with an overall sensitivity of 74.1%, specificity of 60.0% and an ROC of 0.67 (range = 0.538–0.769); 145 dermatologists assessed all 100 clinical images with an overall sensitivity of 89.4%, specificity of 64.4% and an ROC of 0.769 (range = 0.613–0.9). Results between test-sets were significantly different (P < 0.05) confirming the need for a standardised benchmark. Conclusions We present the first public melanoma classification benchmark for both non-dermoscopic and dermoscopic images for comparing artificial intelligence algorithms with diagnostic performance of 145 or 157 dermatologists. Melanoma Classification Benchmark should be considered as a reference standard for white-skinned Western populations in the field of binary algorithmic melanoma classification. Highlights • This paper provides the first open access melanoma classification benchmark for both non-dermoscopic and dermoscopic images. • Algorithms can now be easily compared to the performance of dermatologists in terms of sensitivity, specificity and ROC. • The melanoma benchmark allows comparability between algorithms of different publications and provides a new reference standard. [ABSTRACT FROM AUTHOR]

• This paper explores probabilistic extensions to Deep Learning algorithms and their application on energy time series forecasting. • The methods tested are concrete Dropouts, deep Ensembles, bayesian neural Networks, deep gaussian Processes, and functional neural processes. • The methods are evaluated in terms of calibration, sharpness, and how well they indicate a lack of knowledge (Epistemic Uncertainty). • Good epistemic uncertainty estimates were given using functional neural Processes, deep Ensembles, and deep gaussian processes methods, while concrete dropouts and bayesian neural networks needed tunning for better performance. • The application of the methods was feasible for energy time series forecasting, and they produced good estimates of their confidence in a forecast. Deep Learning methods are widely successful and continue to be applied in new fields, including forecasting for energy time series. However, they lack an indication of their confidence in predictions, which helps evaluate, interpret, and improve the forecasting. This paper explores probabilistic extensions to Deep Learning algorithms and their application on energy time series forecasting. The methods tested are Concrete Dropouts, Deep Ensembles, Bayesian Neural Networks, Deep Gaussian Processes, and Functional Neural Processes. For evaluation, two load forecasting scenarios are considered: 1-hour-ahead and 24-hour-ahead load forecasting. The methods are evaluated in terms of calibration, sharpness, and how well they indicate a lack of knowledge (Epistemic Uncertainty). As a reference, simple Neural Network and Quantile Regression models are used. Also, the ARIMA model and Persistence model are implemented to compare the results. Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and Continuous Ranked Probability Score (CRPS) are computed to measure the error performance. Based on endogenous load time series in Germany, the results showed that Concrete Dropout, Deep Ensembles, and Bayesian Neural Networks performed similarly well and better than the reference methods. They also indicated that the simpler Concrete Dropout and Deep Ensembles have an advantage in both speed and convergence time. Good Epistemic Uncertainty estimates were given using Functional Neural Processes, Deep Ensembles, and Deep Gaussian Processes methods, while Concrete dropouts and Bayesian Neural Networks needed tunning for better performance. The application of the methods was feasible for energy time series forecasting, and they produced good estimates of their confidence in a forecast. [ABSTRACT FROM AUTHOR]

In recent years, wind power has emerged as an important source of renewable energy. When onshore and offshore wind farm regions are connected to the grid for power generation, consistent multi-location short-term wind power predictions are extremely valuable in terms of assuring the power system's safety, sustainability, and economic operation. An abrupt variation in wind power generation influences the efficiency of the regional power grid. This makes accurate short-term forecasting essential for high-level planning and scheduling of power grids. To address the issue, this paper presents two variants of recurrent neural networks (RNN): gated recurrent unit (GRU) and long short-term memory (LSTM) models considering substantially better prediction accuracy to forecast a country-wide (Germany) wind power data for daily (t + 1), and multi-step (t + 3, t + 5 , and t + 12) hours ahead. In addition, wind velocities [m/s] measured at heights of 2, 10, and 50-m (above ground level) are exploited as an essential characteristic among the available input variables and evaluated each feature subset based on four training divisions (80-20%, 70-30%, 60-40%, and 50-50%) and compared the results with ARIMA and SVR approaches in the literature. The findings reveal that the RNN-GRU model not only can achieve higher predicting accuracy but also has a faster learning speed over long sequences. • Multivariate analysis of hourly wind power predictions is carried out. • The prediction model integrated RNN-GRU and RNN-LSTM. • A countrywide dataset for Germany in 2016 is employed for the forecast. • Accuracy and computational efficiency were used to evaluate the performance of proposed time series models. • Comparison with reported models and ablation strategies supports the design of individual modules in RNN-GRU. • RNN-GRU with multi-step/multi-location demonstrated potential for large-scale wind farm applications. [ABSTRACT FROM AUTHOR]

In this paper, we propose a Stack Auto-encoder (SAE)-Driven and Semi-Supervised (SSL)-Based Deep Neural Network (DNN) to extract buildings from relatively low-cost satellite near infrared images. The novelty of our scheme is that we employ only an extremely small portion of labeled data for training the deep model which constitutes less than 0.08% of the total data. This way, we significantly reduce the manual effort needed to complete an annotation process, and thus the time required for creating a reliable labeled dataset. On the contrary, we apply novel semi-supervised techniques to estimate soft labels (targets) of the vast amount of existing unlabeled data and then we utilize these soft estimates to improve model training. Overall, four SSL schemes are employed, the Anchor Graph, the Safe Semi-Supervised Regression (SAFER), the Squared-loss Mutual Information Regularization (SMIR), and an equal importance Weighted Average of them (WeiAve). To retain only the most meaning information of the input data, labeled and unlabeled ones, we also employ a Stack Autoencoder (SAE) trained under an unsupervised manner. This way, we handle noise in the input signals, attributed to dimensionality redundancy, without sacrificing meaningful information. Experimental results on the benchmarked dataset of Vaihingen city in Germany indicate that our approach outperforms all state-of-the-art methods in the field using the same type of color orthoimages, though the fact that a limited dataset is utilized (10 times less data or better, compared to other approaches), while our performance is close to the one achieved by high expensive and much more precise input information like the one derived from Light Detection and Ranging (LiDAR) sensors. In addition, the proposed approach can be easily expanded to handle any number of classes, including buildings, vegetation, and ground. [ABSTRACT FROM AUTHOR]

Storms can cause significant damage to forest areas, affecting biodiversity and infrastructure and leading to economic loss. Thus, rapid detection and mapping of windthrows are crucially important for forest management. Recent advances in computer vision have led to highly-accurate image classification algorithms such as Convolutional Neural Network (CNN) architectures. In this study, we tested and implemented an algorithm based on CNNs in an ArcGIS environment for automatic detection and mapping of damaged areas. The algorithm was trained and tested on a forest area in Bavaria, Germany.. It is a based on a modified U-Net architecture that was optimized for the pixelwise classification of multispectral aerial remote sensing data. The neural network was trained on labeled damaged areas from after-storm aerial orthophotos of a ca. 109 k m 2 forest area with RGB and NIR bands and 0.2-m spatial resolution. Around 10 7 pixels of labeled data were used in the process. Once the network is trained, predictions on further datasets can be computed within seconds, depending on the size of the input raster and the computational power used. The overall accuracy on our test dataset was 92 % . During visual validation, labeling errors were found in the reference data that somewhat biased the results because the algorithm in some instance performed better than the human labeling procedure, while missing areas affected by shadows. Our results are very good in terms of precision, and the methods introduced in this paper have several additional advantages compared to traditional methods: CNNs automatically detect high- and low-level features in the data, leading to high classification accuracies, while only one after-storm image is needed in comparison to two images for approaches based on change detection. Furthermore, flight parameters do not affect the results in the same way as for approaches that require DSMs and DTMs as the classification is only based on the image data themselves, and errors occurring in the computation of DSMs and DTMs do not affect the results with respect to the z component. The integration into the ArcGIS Platform allows a streamlined workflow for forest management, as the results can be accessed by mobile devices in the field to allow for high-accuracy ground-truthing and additional mapping that can be synchronized back into the database. Our results and the provided automatic workflow highlight the potential of deep learning on high-resolution imagery and GIS for fast and efficient post-disaster damage assessment as a first step of disaster management. [ABSTRACT FROM AUTHOR]