Ribana Roscher

Mohamad Hakam Shams Eddin, Ribana Roscher, Juergen Gall

Climate change is expected to intensify and increase extreme events in the weather cycle. Since this has a significant impact on various sectors of our life, recent works are concerned with identifying and predicting such extreme events from Earth observations. With respect to wildfire danger forecasting, previous deep learning approaches duplicate static variables along the time dimension and neglect the intrinsic differences between static and dynamic variables. Furthermore, most existing multi-branch architectures lose the interconnections between the branches during the feature learning stage. To address these issues, this paper proposes a 2D/3D two-branch convolutional neural network (CNN) with a Location-aware Adaptive Normalization layer (LOAN). Using LOAN as a building block, we can modulate the dynamic features conditional on their geographical locations. Thus, our approach considers feature properties as a unified yet compound 2D/3D model. Besides, we propose using the sinusoidal-based encoding of the day of the year to provide the model with explicit temporal information about the target day within the year. Our experimental results show a better performance of our approach than other baselines on the challenging FireCube dataset. The results show that location-aware adaptive feature normalization is a promising technique to learn the relation between dynamic variables and their geographic locations, which is highly relevant for areas where remote sensing data builds the basis for analysis. The source code is available at https://github.com/HakamShams/LOAN.

Jana Kierdorf, Laura Verena Junker-Frohn, Mike Delaney et al.

This article presents GrowliFlower, a georeferenced, image-based UAV time series dataset of two monitored cauliflower fields of size 0.39 and 0.60 ha acquired in 2020 and 2021. The dataset contains RGB and multispectral orthophotos from which about 14,000 individual plant coordinates are derived and provided. The coordinates enable the dataset users the extraction of complete and incomplete time series of image patches showing individual plants. The dataset contains collected phenotypic traits of 740 plants, including the developmental stage as well as plant and cauliflower size. As the harvestable product is completely covered by leaves, plant IDs and coordinates are provided to extract image pairs of plants pre and post defoliation, to facilitate estimations of cauliflower head size. Moreover, the dataset contains pixel-accurate leaf and plant instance segmentations, as well as stem annotations to address tasks like classification, detection, segmentation, instance segmentation, and similar computer vision tasks. The dataset aims to foster the development and evaluation of machine learning approaches. It specifically focuses on the analysis of growth and development of cauliflower and the derivation of phenotypic traits to foster the development of automation in agriculture. Two baseline results of instance segmentation at plant and leaf level based on the labeled instance segmentation data are presented. The entire data set is publicly available.

Ahmed Emam, Mohamed Farag, Ribana Roscher

Protected natural areas are regions that have been minimally affected by human activities such as urbanization, agriculture, and other human interventions. To better understand and map the naturalness of these areas, machine learning models can be used to analyze satellite imagery. Specifically, explainable machine learning methods show promise in uncovering patterns that contribute to the concept of naturalness within these protected environments. Additionally, addressing the uncertainty inherent in machine learning models is crucial for a comprehensive understanding of this concept. However, existing approaches have limitations. They either fail to provide explanations that are both valid and objective or struggle to offer a quantitative metric that accurately measures the contribution of specific patterns to naturalness, along with the associated confidence. In this paper, we propose a novel framework called the Confident Naturalness Explanation (CNE) framework. This framework combines explainable machine learning and uncertainty quantification to assess and explain naturalness. We introduce a new quantitative metric that describes the confident contribution of patterns to the concept of naturalness. Furthermore, we generate an uncertainty-aware segmentation mask for each input sample, highlighting areas where the model lacks knowledge. To demonstrate the effectiveness of our framework, we apply it to a study site in Fennoscandia using two open-source satellite datasets.

Ivica Obadic, Ribana Roscher, Dario Augusto Borges Oliveira et al.

Transformer models have become a promising approach for crop-type classification. Although their attention weights can be used to understand the relevant time points for crop disambiguation, the validity of these insights depends on how closely the attention weights approximate the actual workings of these black-box models, which is not always clear. In this paper, we introduce a novel explainability framework that systematically evaluates the explanatory power of the attention weights of a standard transformer encoder for crop-type classification. Our results show that attention patterns strongly relate to key dates, which are often associated with critical phenological events for crop-type classification. Further, the sensitivity analysis reveals the limited capability of the attention weights to characterize crop phenology as the identified phenological events depend on the other crops considered during training. This limitation highlights the relevance of future work towards the development of deep learning approaches capable of automatically learning the temporal vegetation dynamics for accurate crop disambiguation

Hiba Najjar, Miro Miranda, Marlon Nuske et al.

Crop yield forecasting plays a significant role in addressing growing concerns about food security and guiding decision-making for policymakers and farmers. When deep learning is employed, understanding the learning and decision-making processes of the models, as well as their interaction with the input data, is crucial for establishing trust in the models and gaining insight into their reliability. In this study, we focus on the task of crop yield prediction, specifically for soybean, wheat, and rapeseed crops in Argentina, Uruguay, and Germany. Our goal is to develop and explain predictive models for these crops, using a large dataset of satellite images, additional data modalities, and crop yield maps. We employ a long short-term memory network and investigate the impact of using different temporal samplings of the satellite data and the benefit of adding more relevant modalities. For model explainability, we utilize feature attribution methods to quantify input feature contributions, identify critical growth stages, analyze yield variability at the field level, and explain less accurate predictions. The modeling results show an improvement when adding more modalities or using all available instances of satellite data. The explainability results reveal distinct feature importance patterns for each crop and region. We further found that the most influential growth stages on the prediction are dependent on the temporal sampling of the input data. We demonstrated how these critical growth stages, which hold significant agronomic value, closely align with the existing literature in agronomy and crop development biology.

Ahmed Emam, Timo T. Stomberg, Ribana Roscher

Natural protected areas are vital for biodiversity, climate change mitigation, and supporting ecological processes. Despite their significance, comprehensive mapping is hindered by a lack of understanding of their characteristics and a missing land cover class definition. This paper aims to advance the explanation of the designating patterns forming protected and wild areas. To this end, we propose a novel framework that uses activation maximization and a generative adversarial model. With this, we aim to generate satellite images that, in combination with domain knowledge, are capable of offering complete and valid explanations for the spatial and spectral patterns that define the natural authenticity of these regions. Our proposed framework produces more precise attribution maps pinpointing the designating patterns forming the natural authenticity of protected areas. Our approach fosters our understanding of the ecological integrity of the protected natural areas and may contribute to future monitoring and preservation efforts.

Timo T. Stomberg, Taylor Stone, Johannes Leonhardt et al.

Wilderness areas offer important ecological and social benefits and there are urgent reasons to discover where their positive characteristics and ecological functions are present and able to flourish. We apply a novel explainable machine learning technique to satellite images which show wild and anthropogenic areas in Fennoscandia. Occluding certain activations in an interpretable artificial neural network we complete a comprehensive sensitivity analysis regarding wild and anthropogenic characteristics. This enables us to predict detailed and high-resolution sensitivity maps highlighting these characteristics. Our artificial neural network provides an interpretable activation space increasing confidence in our method. Within the activation space, regions are semantically arranged. Our approach advances explainable machine learning for remote sensing, offers opportunities for comprehensive analyses of existing wilderness, and has practical relevance for conservation efforts.

Burak Ekim, Timo T. Stomberg, Ribana Roscher et al.

Antrophonegic pressure (i.e. human influence) on the environment is one of the largest causes of the loss of biological diversity. Wilderness areas, in contrast, are home to undisturbed ecological processes. However, there is no biophysical definition of the term wilderness. Instead, wilderness is more of a philosophical or cultural concept and thus cannot be easily delineated or categorized in a technical manner. With this paper, (i) we introduce the task of wilderness mapping by means of machine learning applied to satellite imagery (ii) and publish MapInWild, a large-scale benchmark dataset curated for that task. MapInWild is a multi-modal dataset and comprises various geodata acquired and formed from a diverse set of Earth observation sensors. The dataset consists of 8144 images with a shape of 1920 x 1920 pixels and is approximately 350 GB in size. The images are weakly annotated with three classes derived from the World Database of Protected Areas - Strict Nature Reserves, Wilderness Areas, and National Parks. With the dataset, which shall serve as a testbed for developments in fields such as explainable machine learning and environmental remote sensing, we hope to contribute to a deepening of our understanding of the question "What makes nature wild?".

Felix Kröber, Genc Hoxha, Ribana Roscher

Label noise in the sense of incorrect labels is present in many real-world data sets and is known to severely limit the generalizability of deep learning models. In the field of remote sensing, however, automated treatment of label noise in data sets has received little attention to date. In particular, there is a lack of systematic analysis of the performance of data-centric methods that not only cope with label noise but also explicitly identify and isolate noisy labels. In this paper, we examine three such methods and evaluate their behavior under different label noise assumptions. To do this, we inject different types of label noise with noise levels ranging from 10 to 70% into two benchmark data sets, followed by an analysis of how well the selected methods filter the label noise and how this affects task performances. With our analyses, we clearly prove the value of data-centric methods for both parts - label noise identification and task performance improvements. Our analyses provide insights into which method is the best choice depending on the setting and objective. Finally, we show in which areas there is still a need for research in the transfer of data-centric label noise methods to remote sensing data. As such, our work is a step forward in bridging the methodological establishment of data-centric label noise methods and their usage in practical settings in the remote sensing domain.

Keiller Nogueira, Akram Zaytar, Wanli Ma et al.

The increasing accessibility of remotely sensed data and the potential of such data to inform large-scale decision-making has driven the development of deep learning models for many Earth Observation tasks. Traditionally, such models must be trained on large datasets. However, the common assumption that broadly larger datasets lead to better outcomes tends to overlook the complexities of the data distribution, the potential for introducing biases and noise, and the computational resources required for processing and storing vast datasets. Therefore, effective solutions should consider both the quantity and quality of data. In this paper, we propose six novel core-set selection methods for selecting important subsets of samples from remote sensing image segmentation datasets that rely on imagery only, labels only, and a combination of each. We benchmark these approaches against a random-selection baseline on three commonly used land cover classification datasets: DFC2022, Vaihingen, and Potsdam. In each of the datasets, we demonstrate that training on a subset of samples outperforms the random baseline, and some approaches outperform training on all available data. This result shows the importance and potential of data-centric learning for the remote sensing domain. The code is available at https://github.com/keillernogueira/data-centric-rs-classification/.

Ahmed Emam, Mohamed Elbassiouny, Julius Miller et al.

Pollinator insects such as honeybees and bumblebees are vital to global food production and ecosystem stability, yet their populations are declining due to anthropogenic and environmental stressors. Scalable, automated monitoring in agricultural environments remains an open challenge due to the difficulty of detecting small, fast-moving, and often camouflaged insects. To address this, we present BuzzSet v1.0, a large-scale dataset of high-resolution pollinator images collected under real field conditions. BuzzSet contains 7,856 manually verified images with more than 8,000 annotated instances across three classes: honeybees, bumblebees, and unidentified insects. Initial annotations were produced using a YOLOv12 model trained on external data and refined through human verification with open-source tools. All images were preprocessed into 256 x 256 tiles to improve the detection of small insects. We provide baselines using the RF-DETR transformer-based object detector. The model achieves strong classification accuracy with F1 scores of 0.94 and 0.92 for honeybees and bumblebees, with minimal confusion between these categories. The unidentified class remains more difficult due to label ambiguity and fewer samples, yet still contributes insights for robustness evaluation. Overall detection performance (mAP at 0.50 of 0.559) illustrates the challenging nature of the dataset and its potential to drive advances in small object detection under realistic ecological conditions. Future work focuses on expanding the dataset to version 2.0 with additional annotations and evaluating further detection strategies. BuzzSet establishes a benchmark for ecological computer vision, with the primary challenge being reliable detection of insects frequently camouflaged within natural vegetation, highlighting an open problem for future research.

Ribana Roscher, Marc Rußwurm, Caroline Gevaert et al.

Recent developments and research in modern machine learning have led to substantial improvements in the geospatial field. Although numerous deep learning architectures and models have been proposed, the majority of them have been solely developed on benchmark datasets that lack strong real-world relevance. Furthermore, the performance of many methods has already saturated on these datasets. We argue that a shift from a model-centric view to a complementary data-centric perspective is necessary for further improvements in accuracy, generalization ability, and real impact on end-user applications. Furthermore, considering the entire machine learning cycle-from problem definition to model deployment with feedback-is crucial for enhancing machine learning models that can be reliable in unforeseen situations. This work presents a definition as well as a precise categorization and overview of automated data-centric learning approaches for geospatial data. It highlights the complementary role of data-centric learning with respect to model-centric in the larger machine learning deployment cycle. We review papers across the entire geospatial field and categorize them into different groups. A set of representative experiments shows concrete implementation examples. These examples provide concrete steps to act on geospatial data with data-centric machine learning approaches.

Lukas Drees, Dereje T. Demie, Madhuri R. Paul et al.

Image-based crop growth modeling can substantially contribute to precision agriculture by revealing spatial crop development over time, which allows an early and location-specific estimation of relevant future plant traits, such as leaf area or biomass. A prerequisite for realistic and sharp crop image generation is the integration of multiple growth-influencing conditions in a model, such as an image of an initial growth stage, the associated growth time, and further information about the field treatment. We present a two-stage framework consisting first of an image prediction model and second of a growth estimation model, which both are independently trained. The image prediction model is a conditional Wasserstein generative adversarial network (CWGAN). In the generator of this model, conditional batch normalization (CBN) is used to integrate different conditions along with the input image. This allows the model to generate time-varying artificial images dependent on multiple influencing factors of different kinds. These images are used by the second part of the framework for plant phenotyping by deriving plant-specific traits and comparing them with those of non-artificial (real) reference images. For various crop datasets, the framework allows realistic, sharp image predictions with a slight loss of quality from short-term to long-term predictions. Simulations of varying growth-influencing conditions performed with the trained framework provide valuable insights into how such factors relate to crop appearances, which is particularly useful in complex, less explored crop mixture systems. Further results show that adding process-based simulated biomass as a condition increases the accuracy of the derived phenotypic traits from the predicted images. This demonstrates the potential of our framework to serve as an interface between an image- and process-based crop growth model.

Ribana Roscher, Lukas Roth, Cyrill Stachniss et al.

In response to the increasing global demand for food, feed, fiber, and fuel, digital agriculture is rapidly evolving to meet these demands while reducing environmental impact. This evolution involves incorporating data science, machine learning, sensor technologies, robotics, and new management strategies to establish a more sustainable agricultural framework. So far, machine learning research in digital agriculture has predominantly focused on model-centric approaches, focusing on model design and evaluation. These efforts aim to optimize model accuracy and efficiency, often treating data as a static benchmark. Despite the availability of agricultural data and methodological advancements, a saturation point has been reached, with many established machine learning methods achieving comparable levels of accuracy and facing similar limitations. To fully realize the potential of digital agriculture, it is crucial to have a comprehensive understanding of the role of data in the field and to adopt data-centric machine learning. This involves developing strategies to acquire and curate valuable data and implementing effective learning and evaluation strategies that utilize the intrinsic value of data. This approach has the potential to create accurate, generalizable, and adaptable machine learning methods that effectively and sustainably address agricultural tasks such as yield prediction, weed detection, and early disease identification

Johannes Leonhardt, Juergen Gall, Ribana Roscher

Large-scale land cover maps generated using deep learning play a critical role across a wide range of Earth science applications. Open in-situ datasets from principled land cover surveys offer a scalable alternative to manual annotation for training such models. However, their sparse spatial coverage often leads to fragmented and noisy predictions when used with existing deep learning-based land cover mapping approaches. A promising direction to address this issue is object-based classification, which assigns labels to semantically coherent image regions rather than individual pixels, thereby imposing a minimum mapping unit. Despite this potential, object-based methods remain underexplored in deep learning-based land cover mapping pipelines, especially in the context of medium-resolution imagery and sparse supervision. To address this gap, we propose LC-SLab, the first deep learning framework for systematically exploring object-based deep learning methods for large-scale land cover classification under sparse supervision. LC-SLab supports both input-level aggregation via graph neural networks, and output-level aggregation by postprocessing results from established semantic segmentation models. Additionally, we incorporate features from a large pre-trained network to improve performance on small datasets. We evaluate the framework on annual Sentinel-2 composites with sparse LUCAS labels, focusing on the tradeoff between accuracy and fragmentation, as well as sensitivity to dataset size. Our results show that object-based methods can match or exceed the accuracy of common pixel-wise models while producing substantially more coherent maps. Input-level aggregation proves more robust on smaller datasets, whereas output-level aggregation performs best with more data. Several configurations of LC-SLab also outperform existing land cover products, highlighting the framework's practical utility.

Ahmed Emam, Ribana Roscher

Protected natural areas play a vital role in ecological balance and ecosystem services. Monitoring these regions at scale using satellite imagery and machine learning is promising, but current methods often lack interpretability and uncertainty-awareness, and do not address how uncertainty affects naturalness assessment. In contrast, we propose Confidence-Filtered Relevance (CFR), a data-centric framework that combines LRP Attention Rollout with Deep Deterministic Uncertainty (DDU) estimation to analyze how model uncertainty influences the interpretability of relevance heatmaps. CFR partitions the dataset into subsets based on uncertainty thresholds, enabling systematic analysis of how uncertainty shapes the explanations of naturalness in satellite imagery. Applied to the AnthroProtect dataset, CFR assigned higher relevance to shrublands, forests, and wetlands, aligning with other research on naturalness assessment. Moreover, our analysis shows that as uncertainty increases, the interpretability of these relevance heatmaps declines and their entropy grows, indicating less selective and more ambiguous attributions. CFR provides a data-centric approach to assess the relevance of patterns to naturalness in satellite imagery based on their associated certainty.

Jana Kierdorf, Ribana Roscher

Cauliflower is a hand-harvested crop that must fulfill high-quality standards in sales making the timing of harvest important. However, accurately determining harvest-readiness can be challenging due to the cauliflower head being covered by its canopy. While deep learning enables automated harvest-readiness estimation, errors can occur due to field-variability and limited training data. In this paper, we analyze the reliability of a harvest-readiness classifier with interpretable machine learning. By identifying clusters of saliency maps, we derive reliability scores for each classification result using knowledge about the domain and the image properties. For unseen data, the reliability can be used to (i) inform farmers to improve their decision-making and (ii) increase the model prediction accuracy. Using RGB images of single cauliflower plants at different developmental stages from the GrowliFlower dataset, we investigate various saliency mapping approaches and find that they result in different quality of reliability scores. With the most suitable interpretation tool, we adjust the classification result and achieve a 15.72% improvement of the overall accuracy to 88.14% and a 15.44% improvement of the average class accuracy to 88.52% for the GrowliFlower dataset.

Maurice Günder, Facundo R. Ispizua Yamati, Jana Kierdorf et al.

UAV-based image retrieval in modern agriculture enables gathering large amounts of spatially referenced crop image data. In large-scale experiments, however, UAV images suffer from containing a multitudinous amount of crops in a complex canopy architecture. Especially for the observation of temporal effects, this complicates the recognition of individual plants over several images and the extraction of relevant information tremendously. In this work, we present a hands-on workflow for the automatized temporal and spatial identification and individualization of crop images from UAVs abbreviated as "cataloging" based on comprehensible computer vision methods. We evaluate the workflow on two real-world datasets. One dataset is recorded for observation of Cercospora leaf spot - a fungal disease - in sugar beet over an entire growing cycle. The other one deals with harvest prediction of cauliflower plants. The plant catalog is utilized for the extraction of single plant images seen over multiple time points. This gathers large-scale spatio-temporal image dataset that in turn can be applied to train further machine learning models including various data layers. The presented approach improves analysis and interpretation of UAV data in agriculture significantly. By validation with some reference data, our method shows an accuracy that is similar to more complex deep learning-based recognition techniques. Our workflow is able to automatize plant cataloging and training image extraction, especially for large datasets.

Benedikt Mersch, Thomas Höllen, Kun Zhao et al.

The ability to predict the future movements of other vehicles is a subconscious and effortless skill for humans and key to safe autonomous driving. Therefore, trajectory prediction for autonomous cars has gained a lot of attention in recent years. It is, however, still a hard task to achieve human-level performance. Interdependencies between vehicle behaviors and the multimodal nature of future intentions in a dynamic and complex driving environment render trajectory prediction a challenging problem. In this work, we propose a new, data-driven approach for predicting the motion of vehicles in a road environment. The model allows for inferring future intentions from the past interaction among vehicles in highway driving scenarios. Using our neighborhood-based data representation, the proposed system jointly exploits correlations in the spatial and temporal domain using convolutional neural networks. Our system considers multiple possible maneuver intentions and their corresponding motion and predicts the trajectory for five seconds into the future. We implemented our approach and evaluated it on two highway datasets taken in different countries and are able to achieve a competitive prediction performance.

Jakob Gawlikowski, Cedrique Rovile Njieutcheu Tassi, Mohsin Ali et al.

Due to their increasing spread, confidence in neural network predictions became more and more important. However, basic neural networks do not deliver certainty estimates or suffer from over or under confidence. Many researchers have been working on understanding and quantifying uncertainty in a neural network's prediction. As a result, different types and sources of uncertainty have been identified and a variety of approaches to measure and quantify uncertainty in neural networks have been proposed. This work gives a comprehensive overview of uncertainty estimation in neural networks, reviews recent advances in the field, highlights current challenges, and identifies potential research opportunities. It is intended to give anyone interested in uncertainty estimation in neural networks a broad overview and introduction, without presupposing prior knowledge in this field. A comprehensive introduction to the most crucial sources of uncertainty is given and their separation into reducible model uncertainty and not reducible data uncertainty is presented. The modeling of these uncertainties based on deterministic neural networks, Bayesian neural networks, ensemble of neural networks, and test-time data augmentation approaches is introduced and different branches of these fields as well as the latest developments are discussed. For a practical application, we discuss different measures of uncertainty, approaches for the calibration of neural networks and give an overview of existing baselines and implementations. Different examples from the wide spectrum of challenges in different fields give an idea of the needs and challenges regarding uncertainties in practical applications. Additionally, the practical limitations of current methods for mission- and safety-critical real world applications are discussed and an outlook on the next steps towards a broader usage of such methods is given.

Jana Kierdorf, Immanuel Weber, Anna Kicherer et al.

The need for accurate yield estimates for viticulture is becoming more important due to increasing competition in the wine market worldwide. One of the most promising methods to estimate the harvest is berry counting, as it can be approached non-destructively, and its process can be automated. In this article, we present a method that addresses the challenge of occluded berries with leaves to obtain a more accurate estimate of the number of berries that will enable a better estimate of the harvest. We use generative adversarial networks, a deep learning-based approach that generates a likely scenario behind the leaves exploiting learned patterns from images with non-occluded berries. Our experiments show that the estimate of the number of berries after applying our method is closer to the manually counted reference. In contrast to applying a factor to the berry count, our approach better adapts to local conditions by directly involving the appearance of the visible berries. Furthermore, we show that our approach can identify which areas in the image should be changed by adding new berries without explicitly requiring information about hidden areas.

Lukas Drees, Laura Verena Junker-Frohn, Jana Kierdorf et al.

Farmers frequently assess plant growth and performance as basis for making decisions when to take action in the field, such as fertilization, weed control, or harvesting. The prediction of plant growth is a major challenge, as it is affected by numerous and highly variable environmental factors. This paper proposes a novel monitoring approach that comprises high-throughput imaging sensor measurements and their automatic analysis to predict future plant growth. Our approach's core is a novel machine learning-based generative growth model based on conditional generative adversarial networks, which is able to predict the future appearance of individual plants. In experiments with RGB time-series images of laboratory-grown Arabidopsis thaliana and field-grown cauliflower plants, we show that our approach produces realistic, reliable, and reasonable images of future growth stages. The automatic interpretation of the generated images through neural network-based instance segmentation allows the derivation of various phenotypic traits that describe plant growth.

Devis Tuia, Ribana Roscher, Jan Dirk Wegner et al.

In the last years we have witnessed the fields of geosciences and remote sensing and artificial intelligence to become closer. Thanks to both the massive availability of observational data, improved simulations, and algorithmic advances, these disciplines have found common objectives and challenges to advance the modeling and understanding of the Earth system. Despite such great opportunities, we also observed a worrying tendency to remain in disciplinary comfort zones applying recent advances from artificial intelligence on well resolved remote sensing problems. Here we take a position on research directions where we think the interface between these fields will have the most impact and become potential game changers. In our declared agenda for AI on Earth sciences, we aim to inspire researchers, especially the younger generations, to tackle these challenges for a real advance of remote sensing and the geosciences.

Immanuel Weber, Jens Bongartz, Ribana Roscher

Object detection in aerial images is an important task in environmental, economic, and infrastructure-related tasks. One of the most prominent applications is the detection of vehicles, for which deep learning approaches are increasingly used. A major challenge in such approaches is the limited amount of data that arises, for example, when more specialized and rarer vehicles such as agricultural machinery or construction vehicles are to be detected. This lack of data contrasts with the enormous data hunger of deep learning methods in general and object recognition in particular. In this article, we address this issue in the context of the detection of road vehicles in aerial images. To overcome the lack of annotated data, we propose a generative approach that generates top-down images by overlaying artificial vehicles created from 2D CAD drawings on artificial or real backgrounds. Our experiments with a modified RetinaNet object detection network show that adding these images to small real-world datasets significantly improves detection performance. In cases of very limited or even no real-world images, we observe an improvement in average precision of up to 0.70 points. We address the remaining performance gap to real-world datasets by analyzing the effect of the image composition of background and objects and give insights into the importance of background.

Laura Zabawa, Anna Kicherer, Lasse Klingbeil et al.

The extraction of phenotypic traits is often very time and labour intensive. Especially the investigation in viticulture is restricted to an on-site analysis due to the perennial nature of grapevine. Traditionally skilled experts examine small samples and extrapolate the results to a whole plot. Thereby different grapevine varieties and training systems, e.g. vertical shoot positioning (VSP) and semi minimal pruned hedges (SMPH) pose different challenges. In this paper we present an objective framework based on automatic image analysis which works on two different training systems. The images are collected semi automatic by a camera system which is installed in a modified grape harvester. The system produces overlapping images from the sides of the plants. Our framework uses a convolutional neural network to detect single berries in images by performing a semantic segmentation. Each berry is then counted with a connected component algorithm. We compare our results with the Mask-RCNN, a state-of-the-art network for instance segmentation and with a regression approach for counting. The experiments presented in this paper show that we are able to detect green berries in images despite of different training systems. We achieve an accuracy for the berry detection of 94.0% in the VSP and 85.6% in the SMPH.

Ribana Roscher, Bastian Bohn, Marco F. Duarte et al.

Machine learning methods have been remarkably successful for a wide range of application areas in the extraction of essential information from data. An exciting and relatively recent development is the uptake of machine learning in the natural sciences, where the major goal is to obtain novel scientific insights and discoveries from observational or simulated data. A prerequisite for obtaining a scientific outcome is domain knowledge, which is needed to gain explainability, but also to enhance scientific consistency. In this article we review explainable machine learning in view of applications in the natural sciences and discuss three core elements which we identified as relevant in this context: transparency, interpretability, and explainability. With respect to these core elements, we provide a survey of recent scientific works that incorporate machine learning and the way that explainable machine learning is used in combination with domain knowledge from the application areas.

Laura Zabawa, Anna Kicherer, Lasse Klingbeil et al.

Yield estimation and forecasting are of special interest in the field of grapevine breeding and viticulture. The number of harvested berries per plant is strongly correlated with the resulting quality. Therefore, early yield forecasting can enable a focused thinning of berries to ensure a high quality end product. Traditionally yield estimation is done by extrapolating from a small sample size and by utilizing historic data. Moreover, it needs to be carried out by skilled experts with much experience in this field. Berry detection in images offers a cheap, fast and non-invasive alternative to the otherwise time-consuming and subjective on-site analysis by experts. We apply fully convolutional neural networks on images acquired with the Phenoliner, a field phenotyping platform. We count single berries in images to avoid the error-prone detection of grapevine clusters. Clusters are often overlapping and can vary a lot in the size which makes the reliable detection of them difficult. We address especially the detection of white grapes directly in the vineyard. The detection of single berries is formulated as a classification task with three classes, namely 'berry', 'edge' and 'background'. A connected component algorithm is applied to determine the number of berries in one image. We compare the automatically counted number of berries with the manually detected berries in 60 images showing Riesling plants in vertical shoot positioned trellis (VSP) and semi minimal pruned hedges (SMPH). We are able to detect berries correctly within the VSP system with an accuracy of 94.0 \% and for the SMPH system with 85.6 \%.

Katharina Franz, Ribana Roscher, Andres Milioto et al.

Global climate change plays an essential role in our daily life. Mesoscale ocean eddies have a significant impact on global warming, since they affect the ocean dynamics, the energy as well as the mass transports of ocean circulation. From satellite altimetry we can derive high-resolution, global maps containing ocean signals with dominating coherent eddy structures. The aim of this study is the development and evaluation of a deep-learning based approach for the analysis of eddies. In detail, we develop an eddy identification and tracking framework with two different approaches that are mainly based on feature learning with convolutional neural networks. Furthermore, state-of-the-art image processing tools and object tracking methods are used to support the eddy tracking. In contrast to previous methods, our framework is able to learn a representation of the data in which eddies can be detected and tracked in more objective and robust way. We show the detection and tracking results on sea level anomalies (SLA) data from the area of Australia and the East Australia current, and compare our two eddy detection and tracking approaches to identify the most robust and objective method.

Lukas Drees, Ribana Roscher, Susanne Wenzel

The estimation of land cover fractions from remote sensing images is a frequently used indicator of the environmental quality. This paper focuses on the quantification of land cover fractions in an urban area of Berlin, Germany, using simulated hyperspectral EnMAP data with a spatial resolution of 30m$\times$30m. We use constrained sparse representation, where each pixel with unknown surface characteristics is expressed by a weighted linear combination of elementary spectra with known land cover class. We automatically determine the elementary spectra from image reference data using archetypal analysis by simplex volume maximization, and combine it with reversible jump Markov chain Monte Carlo method. In our experiments, the estimation of the automatically derived elementary spectra is compared to the estimation obtained by a manually designed spectral library by means of reconstruction error, mean absolute error of the fraction estimates, sum of fractions, $R^2$, and the number of used elementary spectra. The experiments show that a collection of archetypes can be an adequate and efficient alternative to the manually designed spectral library with respect to the mentioned criteria.

Ribana Roscher, Katja Herzog, Annemarie Kunkel et al.

The berry size is one of the most important fruit traits in grapevine breeding. Non-invasive, image-based phenotyping promises a fast and precise method for the monitoring of the grapevine berry size. In the present study an automated image analyzing framework was developed in order to estimate the size of grapevine berries from images in a high-throughput manner. The framework includes (i) the detection of circular structures which are potentially berries and (ii) the classification of these into the class 'berry' or 'non-berry' by utilizing a conditional random field. The approach used the concept of a one-class classification, since only the target class 'berry' is of interest and needs to be modeled. Moreover, the classification was carried out by using an automated active learning approach, i.e no user interaction is required during the classification process and in addition, the process adapts automatically to changing image conditions, e.g. illumination or berry color. The framework was tested on three datasets consisting in total of 139 images. The images were taken in an experimental vineyard at different stages of grapevine growth according to the BBCH scale. The mean berry size of a plant estimated by the framework correlates with the manually measured berry size by $0.88$.

Anne Braakmann-Folgmann, Ribana Roscher, Susanne Wenzel et al.

Sea level change, one of the most dire impacts of anthropogenic global warming, will affect a large amount of the world's population. However, sea level change is not uniform in time and space, and the skill of conventional prediction methods is limited due to the ocean's internal variabi-lity on timescales from weeks to decades. Here we study the potential of neural network methods which have been used successfully in other applications, but rarely been applied for this task. We develop a combination of a convolutional neural network (CNN) and a recurrent neural network (RNN) to ana-lyse both the spatial and the temporal evolution of sea level and to suggest an independent, accurate method to predict interannual sea level anomalies (SLA). We test our method for the northern and equatorial Pacific Ocean, using gridded altimeter-derived SLA data. We show that the used network designs outperform a simple regression and that adding a CNN improves the skill significantly. The predictions are stable over several years.

Anika Bettge, Ribana Roscher, Susanne Wenzel

This paper addresses the land cover classification task for remote sensing images by deep self-taught learning. Our self-taught learning approach learns suitable feature representations of the input data using sparse representation and undercomplete dictionary learning. We propose a deep learning framework which extracts representations in multiple layers and use the output of the deepest layer as input to a classification algorithm. We evaluate our approach using a multispectral Landsat 5 TM image of a study area in the North of Novo Progresso (South America) and the Zurich Summer Data Set provided by the University of Zurich. Experiments indicate that features learned by a deep self-taught learning framework can be used for classification and improve the results compared to classification results using the original feature representation.

Ron Hagensieker, Ribana Roscher, Johannes Rosentreter et al.

Remote sensing satellite data offer the unique possibility to map land use land cover transformations by providing spatially explicit information. However, detection of short-term processes and land use patterns of high spatial-temporal variability is a challenging task. We present a novel framework using multi-temporal TerraSAR-X data and machine learning techniques, namely Discriminative Markov Random Fields with spatio-temporal priors, and Import Vector Machines, in order to advance the mapping of land cover characterized by short-term changes. Our study region covers a current deforestation frontier in the Brazilian state Pará with land cover dominated by primary forests, different types of pasture land and secondary vegetation, and land use dominated by short-term processes such as slash-and-burn activities. The data set comprises multi-temporal TerraSAR-X imagery acquired over the course of the 2014 dry season, as well as optical data (RapidEye, Landsat) for reference. Results show that land use land cover is reliably mapped, resulting in spatially adjusted overall accuracies of up to $79\%$ in a five class setting, yet limitations for the differentiation of different pasture types remain. The proposed method is applicable on multi-temporal data sets, and constitutes a feasible approach to map land use land cover in regions that are affected by high-frequent temporal changes.

Ribana Roscher, Bernd Uebbing, Jürgen Kusche

Satellite radar altimetry is one of the most powerful techniques for measuring sea surface height variations, with applications ranging from operational oceanography to climate research. Over open oceans, altimeter return waveforms generally correspond to the Brown model, and by inversion, estimated shape parameters provide mean surface height and wind speed. However, in coastal areas or over inland waters, the waveform shape is often distorted by land influence, resulting in peaks or fast decaying trailing edges. As a result, derived sea surface heights are then less accurate and waveforms need to be reprocessed by sophisticated algorithms. To this end, this work suggests a novel Spatio-Temporal Altimetry Retracking (STAR) technique. We show that STAR enables the derivation of sea surface heights over the open ocean as well as over coastal regions of at least the same quality as compared to existing retracking methods, but for a larger number of cycles and thus retaining more useful data. Novel elements of our method are (a) integrating information from spatially and temporally neighboring waveforms through a conditional random field approach, (b) sub-waveform detection, where relevant sub-waveforms are separated from corrupted or non-relevant parts through a sparse representation approach, and (c) identifying the final best set of sea surfaces heights from multiple likely heights using Dijkstra's algorithm. We apply STAR to data from the Jason-1, Jason-2 and Envisat missions for study sites in the Gulf of Trieste, Italy and in the coastal region of the Ganges-Brahmaputra-Meghna estuary, Bangladesh. We compare to several established and recent retracking methods, as well as to tide gauge data. Our experiments suggest that the obtained sea surface heights are significantly less affected by outliers when compared to results obtained by other approaches.

Ribana Roscher, Björn Waske

This paper presents a sparse representation-based classification approach with a novel dictionary construction procedure. By using the constructed dictionary sophisticated prior knowledge about the spatial nature of the image can be integrated. The approach is based on the assumption that each image patch can be factorized into characteristic spatial patterns, also called shapelets, and patch-specific spectral information. A set of shapelets is learned in an unsupervised way and spectral information are embodied by training samples. A combination of shapelets and spectral information are represented in an undercomplete spatial-spectral dictionary for each individual patch, where the elements of the dictionary are linearly combined to a sparse representation of the patch. The patch-based classification is obtained by means of the representation error. Experiments are conducted on three well-known hyperspectral image datasets. They illustrate that our proposed approach shows superior results in comparison to sparse representation-based classifiers that use only limited spatial information and behaves competitively with or better than state-of-the-art classifiers utilizing spatial information and kernelized sparse representation-based classifiers.

Ribana Roscher, Björn Waske, Wolfgang Förstner

In this paper we propose an incremental learning strategy for import vector machines (IVM), which is a sparse kernel logistic regression approach. We use the procedure for the concept of self-training for sequential classification of hyperspectral data. The strategy comprises the inclusion of new training samples to increase the classification accuracy and the deletion of non-informative samples to be memory- and runtime-efficient. Moreover, we update the parameters in the incremental IVM model without re-training from scratch. Therefore, the incremental classifier is able to deal with large data sets. The performance of the IVM in comparison to support vector machines (SVM) is evaluated in terms of accuracy and experiments are conducted to assess the potential of the probabilistic outputs of the IVM. Experimental results demonstrate that the IVM and SVM perform similar in terms of classification accuracy. However, the number of import vectors is significantly lower when compared to the number of support vectors and thus, the computation time during classification can be decreased. Moreover, the probabilities provided by IVM are more reliable, when compared to the probabilistic information, derived from an SVM's output. In addition, the proposed self-training strategy can increase the classification accuracy. Overall, the IVM and the its incremental version is worthwhile for the classification of hyperspectral data.

Beate Franke, Jean-François Plante, Ribana Roscher et al.

The need for new methods to deal with big data is a common theme in most scientific fields, although its definition tends to vary with the context. Statistical ideas are an essential part of this, and as a partial response, a thematic program on statistical inference, learning, and models in big data was held in 2015 in Canada, under the general direction of the Canadian Statistical Sciences Institute, with major funding from, and most activities located at, the Fields Institute for Research in Mathematical Sciences. This paper gives an overview of the topics covered, describing challenges and strategies that seem common to many different areas of application, and including some examples of applications to make these challenges and strategies more concrete.