Publications

Probabilistic Embeddings for Frozen Vision-Language Models: Uncertainty Quantification with Gaussian Process Latent Variable Models
Aishwarya Venkataramanan and Paul Bodesheim and Joachim Denzler.
International Conference on Uncertainty in Artificial Intelligence (UAI). 2025. (accepted)
[bibtex] [abstract]

Using {{Image-based AI}} for Insect Monitoring and Conservation - {{InsectAI COST Action
Tom August and Mario V Balzan and Paul Bodesheim and Gunnar Brehm and Lisette {Cant{\'u}-Salazar} and S{\'i}lvia Castro and Joseph Chipperfield and Guillaume Ghisbain and Alba {Gomez-Segura} and J{\'e}r{\'e}mie Goulnik and Quentin Groom and Laurens Hogeweg and Chantal Huijbers and Andreas Kamilaris and Karolis Kazlauskis and Wouter Koch and Dimitri Korsch and Jo{\~a}o Loureiro and Youri Martin and Angeliki F Martinou and Kent McFarland and Xavier Mestdagh and Denis Michez and Charlie Outhwaite and Luca Pegoraro and Nadja Pernat and Lars B. Pettersson and Pavel Pipek and Cristina Preda and David Rolnick and Tobias Roth and David B. Roy and Helen Roy and Veljo Runnel and Martina Sasic and Dmitry Schigel and Julie Koch Sheard and Cecilie Svenningsen and Heliana Teixeira and Nicolas Titeux and Thomas Tscheulin and Elli Tzirkalli and Marijn {van} {der} {Velde} and Roel {van} {Klink} and Nicolas J Vereecken and Sarah Vray and Toke Thomas H{\o}ye.
Research Ideas and Outcomes. 11: pages e134825. 2025.
[bibtex] [pdf] [abstract]

Do Camera Light Traps for Moths Provide Similar Data as Conventional Funnel Light Traps?
Vivian Holzhauer and Dennis Böttger and Paul Bodesheim and Gunnar Brehm.
bioRxiv. pages 2025.02.06.636905. 2025.
[bibtex] [pdf] [abstract]

Abstract: Insects are of crucial importance for terrestrial ecosystems but many populations decline rapidly. Conventional collecting methods are usually time-consuming, resulting in a low temporal, spatial and taxonomic resolution of data. Automated camera light traps (CLT) allow non-lethal monitoring of species-rich moths (Lepidoptera) and other nocturnal insects, but so far little is known about their performance compared to conventional collecting methods.By observing the behaviour of moths in previous field work, we hypothesized that CLT perform well in moth groups in which species tend to sit down quietly after approaching the lamp (such as Geometridae) but worse in moth groups in which species are persistently active (such as Sphingidae).We tested the performance of two CLTs, equipped with Sony alpha 7II (24 megapixel sensor) cameras that resulted in images with approx. 420 dpi resolution. The study was carried out in a forested area near Bielefeld in NW Germany for 196 nights in a row from April to September 2023, and photos were taken every two minutes during the night. All macromoths recognizable in the photos were identified and counted individually. We directly compared the data from the CLTs with moth samples obtained from conventional funnel light traps (FLTs) during 12 nights which were spread across the flight season.The resulting 420 dpi images from the CLTs allowed reliable species identification of all observed macromoths with only few exceptions due to technical problems. In direct comparison during 12 nights, CLTs recorded 39 species exclusively, FLTs recorded 48 species exclusively, and 53 species were recorded by both methods equally. During the whole sampling period of 196 nights in a row, a total of 225 moth species were recorded by CLTs. CLTs tended to record Geometridae better than FLTs whereas Sphingidae tended to be undersampled by the new method. Familes differed in the length to which they remained on the screen of the CLTs.Our study is the first to systematically compare the methods and it shows that CLTs perform overall very well. Results from CLTs differ to a certain extent from conventional trapping methods because they seem to perform worse in groups with highly active species and perform better in calmer groups like geometrid moths. CLTs are promising devices for insect monitoring since they deliver data with high resolution in time, space and taxonomy. The use of artificial intelligence (AI) for the analysis of images is intended as the next logic step.Competing Interest StatementThe authors have declared no competing interest.

Deep {{Image Clustering}} with {{Model-Agnostic Meta-Learning
Kim Bjerge and Paul Bodesheim and Henrik Karstoft.
International {{Joint Conference}} on {{Computer Vision}}, {{Imaging}} and {{Computer Graphics Theory}} and {{Applications}} ({{VISAPP}}). Pages 286-297. 2025.
[bibtex] [pdf] [abstract]

Adaptive Model Selection for Expanded Post Hoc Debiasing and Mitigating Varying Degrees of Spurious Correlations
Jan Blunk and Paul Bodesheim and Joachim Denzler.
International Conference in Computer Analysis of Images and Patterns (CAIP). 2025. (accepted)
[bibtex]

Determining the Community Composition of Herbaceous Species from Images using Convolutional Neural Networks
Matthias Körschens and Solveig Franziska Bucher and Paul Bodesheim and Josephine Ulrich and Joachim Denzler and Christine Römermann.
Ecological Informatics. 80: pages 102516. 2024.
[bibtex] [pdf] [web] [abstract]

Few-{{Shot Learning}} with {{Novelty Detection
Kim Bjerge and Paul Bodesheim and Henrik Karstoft.
International {{Conference}} on {{Deep Learning Theory}} and {{Applications}} ({{DeLTA}}). Pages 340-363. 2024. Best Paper Award
[bibtex] [pdf] [web] [abstract]

Beyond {{Debiasing}}: {{Actively Steering Feature Selection}} via {{Loss Regularization
Jan Blunk and Niklas Penzel and Paul Bodesheim and Joachim Denzler.
DAGM German Conference}} on {{Pattern Recognition}} ({{DAGM-GCPR}}). Pages 394-408. 2023.
[bibtex] [pdf] [abstract]

Beyond Global Average Pooling: Alternative Feature Aggregations for Weakly Supervised Localization
Matthias Körschens and Paul Bodesheim and Joachim Denzler.
International Conference on Computer Vision Theory and Applications (VISAPP). Pages 180-191. 2022.
[bibtex] [pdf] [abstract]

Automated Visual Monitoring of Nocturnal Insects with Light-based Camera Traps
Dimitri Korsch and Paul Bodesheim and Gunnar Brehm and Joachim Denzler.
CVPR Workshop on Fine-grained Visual Classification (CVPR-WS). 2022.
[bibtex] [pdf] [web] [code] [abstract]

Predicting spatiotemporal variability in radial tree growth at the continental scale with machine learning
Paul Bodesheim and Flurin Babst and David C. Frank and Claudia Hartl and Christian S. Zang and Martin Jung and Markus Reichstein and Miguel D. Mahecha.
Environmental Data Science. 1: pages E9. 2022.
[bibtex] [pdf] [abstract]

Pre-trained models are not enough: active and lifelong learning is important for long-term visual monitoring of mammals in biodiversity research. Individual identification and attribute prediction with image features from deep neural networks and decoupled decision models applied to elephants and great apes
Paul Bodesheim and Jan Blunk and Matthias Körschens and Clemens-Alexander Brust and Christoph Käding and Joachim Denzler.
Mammalian Biology. 102: pages 875-897. 2022.
[bibtex] [pdf] [web] [abstract]

Abstract: Animal re-identification based on image data, either recorded manually by photographers or automatically with camera traps, is an important task for ecological studies about biodiversity and conservation that can be highly automatized with algorithms from computer vision and machine learning. However, fixed identification models only trained with standard datasets before their application will quickly reach their limits, especially for long-term monitoring with changing environmental conditions, varying visual appearances of individuals over time that differ a lot from those in the training data, and new occurring individuals that have not been observed before. Hence, we believe that active learning with human-in-the-loop and continuous lifelong learning is important to tackle these challenges and to obtain high-performance recognition systems when dealing with huge amounts of additional data that become available during the application. Our general approach with image features from deep neural networks and decoupled decision models can be applied to many different mammalian species and is perfectly suited for continuous improvements of the recognition systems via lifelong learning. In our identification experiments, we consider four different taxa, namely two elephant species: African forest elephants and Asian elephants, as well as two species of great apes: gorillas and chimpanzees. Going beyond classical re-identification, our decoupled approach can also be used for predicting attributes of individuals such as gender or age using classification or regression methods. Although applicable for small datasets of individuals as well, we argue that even better recognition performance will be achieved by improving decision models gradually via lifelong learning to exploit huge datasets and continuous recordings from long-term applications. We highlight that algorithms for deploying lifelong learning in real observational studies exist and are ready for use. Hence, lifelong learning might become a valuable concept that supports practitioners when analyzing large-scale image data during long-term monitoring of mammals.

Investigating Neural Network Training on a Feature Level using Conditional Independence
Niklas Penzel and Christian Reimers and Paul Bodesheim and Joachim Denzler.
ECCV Workshop on Causality in Vision (ECCV-WS). Pages 383-399. 2022.
[bibtex] [pdf] [abstract]

Emerging technologies revolutionise insect ecology and monitoring
Roel van Klink and Tom August and Yves Bas and Paul Bodesheim and Aletta Bonn and Frode Fossoy and Toke T. Hoye and Eelke Jongejans and Myles H.M. Menz and Andreia Miraldo and Tomas Roslin and Helen E. Roy and Ireneusz Ruczynski and Dmitry Schigel and Livia Schäffler and Julie K. Sheard and Cecilie Svenningsen and Georg F. Tschan and Jana Wäldchen and Vera M.A. Zizka and Jens Aström and Diana E. Bowler.
Trends in Ecology \& Evolution. 37 (10): pages 872-885. 2022.
[bibtex] [pdf] [abstract]

Towards a multisensor station for automated biodiversity monitoring
J. Wolfgang Wägele and Paul Bodesheim and Sarah J. Bourlat and Joachim Denzler and Michael Diepenbroek and Vera Fonseca and Karl-Heinz Frommolt and Matthias F. Geiger and Birgit Gemeinholzer and Frank Oliver Glöckner and Timm Haucke and Ameli Kirse and Alexander Kölpin and Ivaylo Kostadinov and Hjalmar S. Kühl and Frank Kurth and Mario Lasseck and Sascha Liedke and Florian Losch and Sandra Müller and Natalia Petrovskaya and Krzysztof Piotrowski and Bernd Radig and Christoph Scherber and Lukas Schoppmann and Jan Schulz and Volker Steinhage and Georg F. Tschan and Wolfgang Vautz and Domenico Velotto and Maximilian Weigend and Stefan Wildermann.
Basic and Applied Ecology. 59: pages 105-138. 2022.
[bibtex] [pdf] [web] [abstract]

Abstract: Rapid changes of the biosphere observed in recent years are caused by both small and large scale drivers, like shifts in temperature, transformations in land-use, or changes in the energy budget of systems. While the latter processes are easily quantifiable, documentation of the loss of biodiversity and community structure is more difficult. Changes in organismal abundance and diversity are barely documented. Censuses of species are usually fragmentary and inferred by often spatially, temporally and ecologically unsatisfactory simple species lists for individual study sites. Thus, detrimental global processes and their drivers often remain unrevealed. A major impediment to monitoring species diversity is the lack of human taxonomic expertise that is implicitly required for large-scale and fine-grained assessments. Another is the large amount of personnel and associated costs needed to cover large scales, or the inaccessibility of remote but nonetheless affected areas. To overcome these limitations we propose a network of Automated Multisensor stations for Monitoring of species Diversity (AMMODs) to pave the way for a new generation of biodiversity assessment centers. This network combines cutting-edge technologies with biodiversity informatics and expert systems that conserve expert knowledge. Each AMMOD station combines autonomous samplers for insects, pollen and spores, audio recorders for vocalizing animals, sensors for volatile organic compounds emitted by plants (pVOCs) and camera traps for mammals and small invertebrates. AMMODs are largely self-containing and have the ability to pre-process data (e.g. for noise filtering) prior to transmission to receiver stations for storage, integration and analyses. Installation on sites that are difficult to access require a sophisticated and challenging system design with optimum balance between power requirements, bandwidth for data transmission, required service, and operation under all environmental conditions for years. An important prerequisite for automated species identification are databases of DNA barcodes, animal sounds, for pVOCs, and images used as training data for automated species identification. AMMOD stations thus become a key component to advance the field of biodiversity monitoring for research and policy by delivering biodiversity data at an unprecedented spatial and temporal resolution.

Occlusion-Robustness of Convolutional Neural Networks via Inverted Cutout
Matthias Körschens and Paul Bodesheim and Joachim Denzler.
International Conference on Pattern Recognition (ICPR). Pages 2829-2835. 2022.
[bibtex] [pdf] [supplementary] [abstract]

Deep Learning Pipeline for Automated Visual Moth Monitoring: Insect Localization and Species Classification
Dimitri Korsch and Paul Bodesheim and Joachim Denzler.
INFORMATIK 2021, Computer Science for Biodiversity Workshop (CS4Biodiversity). Pages 443-460. 2021.
[bibtex] [pdf] [web] [code] [abstract]

Exploiting Web Images for Moth Species Classification
Julia Böhlke and Dimitri Korsch and Paul Bodesheim and Joachim Denzler.
Computer Science for Biodiversity Workshop (CS4Biodiversity), INFORMATIK 2021. Pages 481-498. 2021.
[bibtex] [pdf] [web] [abstract]

End-to-end Learning of Fisher Vector Encodings for Part Features in Fine-grained Recognition
Dimitri Korsch and Paul Bodesheim and Joachim Denzler.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 142-158. 2021.
[bibtex] [pdf] [web] [code] [abstract]

Conditional Adversarial Debiasing: Towards Learning Unbiased Classifiers from Biased Data
Christian Reimers and Paul Bodesheim and Jakob Runge and Joachim Denzler.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 48-62. 2021.
[bibtex] [pdf] [abstract]

Conditional Dependence Tests Reveal the Usage of ABCD Rule Features and Bias Variables in Automatic Skin Lesion Classification
Christian Reimers and Niklas Penzel and Paul Bodesheim and Jakob Runge and Joachim Denzler.
CVPR ISIC Skin Image Analysis Workshop (CVPR-WS). Pages 1810-1819. 2021.
[bibtex] [pdf] [web] [abstract]

Minimizing the Annotation Effort for Detecting Wildlife in Camera Trap Images with Active Learning
Daphne Auer and Paul Bodesheim and Christian Fiderer and Marco Heurich and Joachim Denzler.
Computer Science for Biodiversity Workshop (CS4Biodiversity), INFORMATIK 2021. Pages 547-564. 2021.
[bibtex] [pdf] [abstract]

Lightweight Filtering of Noisy Web Data: Augmenting Fine-grained Datasets with Selected Internet Images
Julia Böhlke and Dimitri Korsch and Paul Bodesheim and Joachim Denzler.
International Conference on Computer Vision Theory and Applications (VISAPP). Pages 466-477. 2021.
[bibtex] [pdf] [web] [abstract]

Automatic Plant Cover Estimation with Convolutional Neural Networks
Matthias Körschens and Paul Bodesheim and Christine Römermann and Solveig Franziska Bucher and Mirco Migliavacca and Josephine Ulrich and Joachim Denzler.
Computer Science for Biodiversity Workshop (CS4Biodiversity), INFORMATIK 2021. Pages 499-516. 2021.
[bibtex] [pdf] [abstract]

Automated Visual Large Scale Monitoring of Faunal Biodiversity
Bernd Radig and Paul Bodesheim and Dimitri Korsch and Joachim Denzler and Timm Haucke and Morris Klasen and Volker Steinhage.
Pattern Recognition and Image Analysis. Advances in Mathematical Theory and Applications (PRIA). 31 (3): pages 477-488. 2021.
[bibtex] [pdf] [web] [abstract]

Abstract: To observe biodiversity, the variety of plant and animal life in the world or in a particular habitat, human observers make the most common examinations, often assisted by technical equipment. Measuring objectively the number of different species of animals, plants, fungi, and microbes that make up the ecosystem can be difficult. In order to monitor changes in biodiversity, data have to be compared across space and time. Cameras are an essential sensor to determine the species range, abundance, and behavior of animals. The millions of recordings from camera traps set up in natural environments can no longer be analyzed by biologists. We started research on doing this analysis automatically without human interaction. The focus of our present sensor is on image capture of wildlife and moths. Special hardware elements for the detection of different species are designed, implemented, tested, and improved, as well as the algorithms for classification and counting of samples from images and image sequences, e.g., to calculate presence, absence, and abundance values or the duration of characteristic activities related to the spatial mobilities. For this purpose, we are developing stereo camera traps that allow spatial reconstruction of the observed animals. This allows three-dimensional coordinates to be recorded and the shape to be characterized. With this additional feature data, species identification and movement detection are facilitated. To classify and count moths, they are attracted to an illuminated screen, which is then photographed at intervals by a high-resolution color camera. To greatly reduce the volume of data, redundant elements and elements that are consistent from image to image are eliminated. All design decisions take into account that at remote sites and in fully autonomous operation, power supply on the one hand and possibilities for data exchange with central servers on the other hand are limited. Installation at hard-to-reach locations requires a sophisticated and demanding system design with an optimal balance between power requirements, bandwidth for data transmission, required service and operation in all environmental conditions for at least ten years.

Weakly Supervised Segmentation Pretraining for Plant Cover Prediction
Matthias Körschens and Paul Bodesheim and Christine Römermann and Solveig Franziska Bucher and Mirco Migliavacca and Josephine Ulrich and Joachim Denzler.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 589-603. 2021.
[bibtex] [pdf] [supplementary] [abstract]

Towards Confirmable Automated Plant Cover Determination
Matthias Körschens and Paul Bodesheim and Christine Römermann and Solveig Franziska Bucher and Josephine Ulrich and Joachim Denzler.
ECCV Workshop on Computer Vision Problems in Plant Phenotyping (CVPPP). 2020.
[bibtex] [pdf] [web] [supplementary] [abstract]

Earth system data cubes unravel global multivariate dynamics
Miguel D. Mahecha and Fabian Gans and Gunnar Brandt and Rune Christiansen and Sarah E. Cornell and Normann Fomferra and Guido Kraemer and Jonas Peters and Paul Bodesheim and Gustau Camps-Valls and Jonathan F. Donges and Wouter Dorigo and Lina M. Estupinan-Suarez and Victor H. Gutierrez-Velez and Martin Gutwin and Martin Jung and Maria C. Londono and Diego Miralles and Phillip Papastefanou and Markus Reichstein.
Earth System Dynamics. 11 (1): pages 201-234. 2020.
[bibtex] [pdf] [web] [abstract]

Scaling carbon fluxes from eddy covariance sites to globe: synthesis and evaluation of the FLUXCOM approach
Martin Jung and Christopher Schwalm and Mirco Migliavacca and Sophia Walther and Gustau Camps-Valls and Sujan Koirala and Peter Anthoni and Simon Besnard and Paul Bodesheim and Nuno Carvalhais and Frederic Chevallier and Fabian Gans and Daniel S. Goll and Vanessa Haverd and Philipp Köhler and Kazuhito Ichii and Atul K. Jain and Junzhi Liu and Danica Lombardozzi and Julia E.M.S. Nabel and Jacob A. Nelson and Michael O'Sullivan and Martijn Pallandt and Dario Papale and Wouter Peters and Julia Pongratz and Christian Rödenbeck and Stephen Sitch and Gianluca Tramontana and Anthony Walker and Ulrich Weber and Markus Reichstein.
Biogeosciences. 17 (5): pages 1343-1365. 2020.
[bibtex] [pdf] [web] [abstract]

Abstract: FLUXNET comprises globally distributed eddy-covariance-based estimates of carbon fluxes between the biosphere and the atmosphere. Since eddy covariance flux towers have a relatively small footprint and are distributed unevenly across the world, upscaling the observations is necessary to obtain global-scale estimates of biosphereatmosphere exchange. Based on cross-consistency checks with atmospheric inversions, sun-induced fluorescence (SIF) and dynamic global vegetation models (DGVMs), here we provide a systematic assessment of the latest upscaling efforts for gross primary production (GPP) and net ecosystem exchange (NEE) of the FLUXCOM initiative, where different machine learning methods, forcing data sets and sets of predictor variables were employed. Spatial patterns of mean GPP are consistent across FLUXCOM and DGVM ensembles (R2>0.94 at 1\textdegree{} spatial resolution) while the majority of DGVMs show, for 70\hspace{0.167em}\% of the land surface, values outside the FLUXCOM range. Global mean GPP magnitudes for 20082010 from FLUXCOM members vary within 106 and 130\hspace{0.167em}PgC\hspace{0.167em}yr-1 with the largest uncertainty in the tropics. Seasonal variations in independent SIF estimates agree better with FLUXCOM GPP (mean global pixel-wise R2\sim{}0.75) than with GPP from DGVMs (mean global pixel-wise R2\sim{}0.6). Seasonal variations in FLUXCOM NEE show good consistency with atmospheric inversion-based net land carbon fluxes, particularly for temperate and boreal regions (R2>0.92). Interannual variability of global NEE in FLUXCOM is underestimated compared to inversions and DGVMs. The FLUXCOM version which also uses meteorological inputs shows a strong co-variation in interannual patterns with inversions (R2=0.87 for 20012010). Mean regional NEE from FLUXCOM shows larger uptake than inversion and DGVM-based estimates, particularly in the tropics with discrepancies of up to several hundred grammes of carbon per square metre per year. These discrepancies can only partly be reconciled by carbon loss pathways that are implicit in inversions but not captured by the flux tower measurements such as carbon emissions from fires and water bodies. We hypothesize that a combination of systematic biases in the underlying eddy covariance data, in particular in tall tropical forests, and a lack of site history effects on NEE in FLUXCOM are likely responsible for the too strong tropical carbon sink estimated by FLUXCOM. Furthermore, as FLUXCOM does not account for CO2 fertilization effects, carbon flux trends are not realistic. Overall, current FLUXCOM estimates of mean annual and seasonal cycles of GPP as well as seasonal NEE variations provide useful constraints of global carbon cycling, while interannual variability patterns from FLUXCOM are valuable but require cautious interpretation. Exploring the diversity of Earth observation data and of machine learning concepts along with improved quality and quantity of flux tower measurements will facilitate further improvements of the FLUXCOM approach overall.

Classification-Specific Parts for Improving Fine-Grained Visual Categorization
Dimitri Korsch and Paul Bodesheim and Joachim Denzler.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 62-75. 2019.
[bibtex] [pdf] [web] [code] [abstract]

Upscaled diurnal cycles of land-atmosphere fluxes: a new global half-hourly data product
Paul Bodesheim and Martin Jung and Fabian Gans and Miguel D. Mahecha and Markus Reichstein.
Earth System Science Data. 10 (3): pages 1327-1365. 2018.
[bibtex] [pdf] [web] [abstract]

Abstract: Interactions between the biosphere and the atmosphere can be well characterized by fluxes between the two. In particular, carbon and energy fluxes play a major role in understanding biogeochemical processes on an ecosystem level or global scale. However, the fluxes can only be measured at individual sites, e.g., by eddy covariance towers, and an upscaling of these local observations is required to analyze global patterns. Previous work focused on upscaling monthly, 8-day, or daily average values, and global maps for each flux have been provided accordingly. In this paper, we raise the upscaling of carbon and energy fluxes between land and atmosphere to the next level by increasing the temporal resolution to subdaily timescales. We provide continuous half-hourly fluxes for the period from 2001 to 2014 at 0.5\textdegree{} spatial resolution, which allows for analyzing diurnal cycles globally. The data set contains four fluxes: gross primary production (GPP), net ecosystem exchange (NEE), latent heat (LE), and sensible heat (H). We propose two prediction approaches for the diurnal cycles based on large-scale regression models and compare them in extensive cross-validation experiments using different sets of predictor variables. We analyze the results for a set of FLUXNET tower sites showing the suitability of our approaches for this upscaling task. Finally, we have selected one approach to calculate the global half-hourly data products based on predictor variables from remote sensing and meteorology at daily resolution as well as half-hourly potential radiation. In addition, we provide a derived product that only contains monthly average diurnal cycles, which is a lightweight version in terms of data storage that still allows studying the important characteristics of diurnal patterns globally. We recommend to primarily use these monthly average diurnal cycles, because they are less affected by the impacts of day-to-day variation, observation noise, and short-term fluctuations on subdaily timescales compared to the full half-hourly flux products. The global half-hourly data products are available at <a href="https://doi.org/10.17871/BACI.224">https://doi.org/10.17871/BACI.224</a>.

When tree rings go global: challenges and opportunities for retro-and prospective insight
Flurin Babst and Paul Bodesheim and Noah Charney and Andrew D. Friend and Martin P. Girardin and Stefan Klesse and David J.P. Moore and Kristina Seftigen and Jesper Björklund and Olivier Bouriaud and Andria Dawson and R. Justin DeRose and Michael C. Dietze and Annemarie H. Eckes and Brian Enquist and David C. Frank and Miguel D. Mahecha and Benjamin Poulter and Sydne Record and Valerie Trouet and Rachael H. Turton and Zhen Zhang and Margaret E.K. Evans.
Quaternary Science Reviews. 197: pages 1-20. 2018.
[bibtex] [pdf] [web] [abstract]

Discovering unknown visual objects with novelty detection techniques
Paul Bodesheim. 2018. Dissertation. ISBN: 978-3-8439-3546-3
[bibtex] [web] [abstract]

SMOS Vegetation Optical Depth and Ecosystem Functional Properties: Exploring Their Relationships in Tropical Forests
Gaia Vaglio Laurin and Cristina Vittucci and Gianluca Tramontana and Paul Bodesheim and Paolo Ferrazzoli and Leila Guerriero and Martin Jung and Miguel D. Mahecha and Dario Papale.
IEEE International Geoscience and Remote Sensing Symposium (IGARSS). Pages 5891-5894. 2018.
[bibtex] [pdf] [web] [abstract]

Upscaling diurnal cycles of carbon fluxes
Paul Bodesheim and Martin Jung and Miguel D. Mahecha and Markus Reichstein.
European Geosciences Union General Assembly (EGU): Abstract + Poster Presentation. 2017.
[bibtex] [pdf] [web] [abstract]

Improved tree-ring archives will support earth-system science
Flurin Babst and Benjamin Poulter and Paul Bodesheim and Miguel D. Mahecha and David C. Frank.
Nature Ecology \& Evolution. 1 (2): 2017.
[bibtex] [pdf] [web] [abstract]

When tree rings go global: challenges and opportunities for retro- andprospective insights
Flurin Babst and Olivier Bouriaud and Benjamin Poulter and Zhen Zhang and Valerie Trouet and Margaret Evans and Noah Charney and Sydne Record and Brian Enquist and Kristina Seftigen and Jesper Björklund and Stefan Klesse and Paul Bodesheim and Miguel Mahecha and Martin Girardin and Andrew Friend and David Frank.
European Geosciences Union General Assembly (EGU): Abstract + Oral Presentation. 2017.
[bibtex] [pdf] [web] [abstract]

Multivariate anomaly detection for Earth observations: a comparison of algorithms and feature extraction techniques
Milan Flach and Fabian Gans and Alexander Brenning and Joachim Denzler and Markus Reichstein and Erik Rodner and Sebastian Bathiany and Paul Bodesheim and Yanira Guanche and Sebasitan Sippel and Miguel D. Mahecha.
Earth System Dynamics. 8 (3): pages 677-696. 2017.
[bibtex] [pdf] [web] [abstract]

Abstract: Today, many processes at the Earth's surface are constantly monitored by multiple data streams. These observations have become central to advancing our understanding of vegetation dynamics in response to climate or land use change. Another set of important applications is monitoring effects of extreme climatic events, other disturbances such as fires, or abrupt land transitions. One important methodological question is how to reliably detect anomalies in an automated and generic way within multivariate data streams, which typically vary seasonally and are interconnected across variables. Although many algorithms have been proposed for detecting anomalies in multivariate data, only a few have been investigated in the context of Earth system science applications. In this study, we systematically combine and compare feature extraction and anomaly detection algorithms for detecting anomalous events. Our aim is to identify suitable workflows for automatically detecting anomalous patterns in multivariate Earth system data streams. We rely on artificial data that mimic typical properties and anomalies in multivariate spatiotemporal Earth observations like sudden changes in basic characteristics of time series such as the sample mean, the variance, changes in the cycle amplitude, and trends. This artificial experiment is needed as there is no "gold standard" for the identification of anomalies in real Earth observations. Our results show that a well-chosen feature extraction step (e.g., subtracting seasonal cycles, or dimensionality reduction) is more important than the choice of a particular anomaly detection algorithm. Nevertheless, we identify three detection algorithms (k-nearest neighbors mean distance, kernel density estimation, a recurrence approach) and their combinations (ensembles) that outperform other multivariate approaches as well as univariate extreme-event detection methods. Our results therefore provide an effective workflow to automatically detect anomalies in Earth system science data.

Large-Scale Gaussian Process Inference with Generalized Histogram Intersection Kernels for Visual Recognition Tasks
Erik Rodner and Alexander Freytag and Paul Bodesheim and Björn Fröhlich and Joachim Denzler.
International Journal of Computer Vision (IJCV). 121 (2): pages 253-280. 2017.
[bibtex] [pdf] [web] [abstract]

Potential of new machine learning methods for understanding long-term interannual variability of carbon and energy fluxes and states from site to global scale
Markus Reichstein and Martin Jung and Paul Bodesheim and Miguel D. Mahecha and Fabian Gans and Erik Rodner and Gustau Camps-Valls and Dario Papale and Gianluca Tramontana and Joachim Denzler and Dennis D. Baldocchi.
American Geophysical Union Fall Meeting (AGU): Abstract + Oral Presentation. 2016.
[bibtex] [web] [abstract]

Diagnosing the Dynamics of Observed and Simulated Ecosystem Gross Primary Productivity with Time Causal Information Theory Quantifiers
Sebastian Sippel and Holger Lange and Miguel D. Mahecha and Michael Hauhs and Paul Bodesheim and Thomas Kaminski and Fabian Gans and Osvaldo A. Rosso.
PLOS ONE. 11 (10): pages 1-29. 2016.
[bibtex] [pdf] [web] [abstract]

Abstract: Data analysis and model-data comparisons in the environmental sciences require diagnostic measures that quantify time series dynamics and structure, and are robust to noise in observational data. This paper investigates the temporal dynamics of environmental time series using measures quantifying their information content and complexity. The measures are used to classify natural processes on one hand, and to compare models with observations on the other. The present analysis focuses on the global carbon cycle as an area of research in which model-data integration and comparisons are key to improving our understanding of natural phenomena. We investigate the dynamics of observed and simulated time series of Gross Primary Productivity (GPP), a key variable in terrestrial ecosystems that quantifies ecosystem carbon uptake. However, the dynamics, patterns and magnitudes of GPP time series, both observed and simulated, vary substantially on different temporal and spatial scales. We demonstrate here that information content and complexity, or Information Theory Quantifiers (ITQ) for short, serve as robust and efficient data-analytical and model benchmarking tools for evaluating the temporal structure and dynamical properties of simulated or observed time series at various spatial scales. At continental scale, we compare GPP time series simulated with two models and an observations-based product. This analysis reveals qualitative differences between model evaluation based on ITQ compared to traditional model performance metrics, indicating that good model performance in terms of absolute or relative error does not imply that the dynamics of the observations is captured well. Furthermore, we show, using an ensemble of site-scale measurements obtained from the FLUXNET archive in the Mediterranean, that model-data or model-model mismatches as indicated by ITQ can be attributed to and interpreted as differences in the temporal structure of the respective ecological time series. At global scale, our understanding of C fluxes relies on the use of consistently applied land models. Here, we use ITQ to evaluate model structure: The measures are largely insensitive to climatic scenarios, land use and atmospheric gas concentrations used to drive them, but clearly separate the structure of 13 different land models taken from the CMIP5 archive and an observations-based product. In conclusion, diagnostic measures of this kind provide data-analytical tools that distinguish different types of natural processes based solely on their dynamics, and are thus highly suitable for environmental science applications such as model structural diagnostics.

Maximally Divergent Intervals for Anomaly Detection
Erik Rodner and Björn Barz and Yanira Guanche and Milan Flach and Miguel Mahecha and Paul Bodesheim and Markus Reichstein and Joachim Denzler.
Workshop on Anomaly Detection (ICML-WS). 2016. Best Paper Award
[bibtex] [pdf] [web] [code] [abstract]

Using Statistical Process Control for detecting anomalies in multivariate spatiotemporal Earth Observations
Milan Flach and Miguel Mahecha and Fabian Gans and Erik Rodner and Paul Bodesheim and Yanira Guanche-Garcia and Alexander Brenning and Joachim Denzler and Markus Reichstein.
European Geosciences Union General Assembly (EGU): Abstract + Oral Presentation. 2016.
[bibtex] [pdf] [web] [abstract]

Abstract: The number of available Earth observations (EOs) is currently substantially increasing. Detecting anomalous pat-terns in these multivariate time series is an important step in identifying changes in the underlying dynamicalsystem. Likewise, data quality issues might result in anomalous multivariate data constellations and have to beidentified before corrupting subsequent analyses. In industrial application a common strategy is to monitor pro-duction chains with several sensors coupled to some statistical process control (SPC) algorithm. The basic ideais to raise an alarm when these sensor data depict some anomalous pattern according to the SPC, i.e. the produc-tion chain is considered 'out of control'. In fact, the industrial applications are conceptually similar to the on-linemonitoring of EOs. However, algorithms used in the context of SPC or process monitoring are rarely consideredfor supervising multivariate spatio-temporal Earth observations. The objective of this study is to exploit the poten-tial and transferability of SPC concepts to Earth system applications. We compare a range of different algorithmstypically applied by SPC systems and evaluate their capability to detect e.g. known extreme events in land sur-face processes. Specifically two main issues are addressed: (1) identifying the most suitable combination of datapre-processing and detection algorithm for a specific type of event and (2) analyzing the limits of the individual ap-proaches with respect to the magnitude, spatio-temporal size of the event as well as the data's signal to noise ratio.Extensive artificial data sets that represent the typical properties of Earth observations are used in this study. Ourresults show that the majority of the algorithms used can be considered for the detection of multivariate spatiotem-poral events and directly transferred to real Earth observation data as currently assembled in different projectsat the European scale, e.g. http://baci-h2020.eu/index.php/ and http://earthsystemdatacube.net/. Known anomaliessuch as the Russian heatwave are detected as well as anomalies which are not detectable with univariate methods.

Hot spots of multivariate extreme anomalies in Earth observations
Milan Flach and Sebastian Sippel and Paul Bodesheim and Alexander Brenning and Joachim Denzler and Fabian Gans and Yanira Guanche and Markus Reichstein and Erik Rodner and Miguel D. Mahecha.
American Geophysical Union Fall Meeting (AGU): Abstract + Oral Presentation. 2016.
[bibtex] [web] [abstract]

Ranking drivers of global carbon and energy fluxes over land
Gustau Camps-Valls and Martin Jung and Kazuhito Ichii and Dario Papale and Gianluca Tramontana and Paul Bodesheim and Christopher Schwalm and Jakob Zscheischler and Miguel D. Mahecha and Markus Reichstein.
IEEE International Geoscience and Remote Sensing Symposium (IGARSS). Pages 4416-4419. 2015.
[bibtex] [pdf] [web] [abstract]

Active Learning and Discovery of Object Categories in the Presence of Unnameable Instances
Christoph Käding and Alexander Freytag and Erik Rodner and Paul Bodesheim and Joachim Denzler.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Pages 4343-4352. 2015.
[bibtex] [pdf] [web] [code] [presentation] [supplementary] [abstract]

Local Novelty Detection in Multi-class Recognition Problems
Paul Bodesheim and Alexander Freytag and Erik Rodner and Joachim Denzler.
IEEE Winter Conference on Applications of Computer Vision (WACV). Pages 813-820. 2015.
[bibtex] [pdf] [web] [supplementary] [abstract]

Temporal Video Segmentation by Event Detection: A Novelty Detection Approach
Mahesh Venkata Krishna and Paul Bodesheim and Marco Körner and Joachim Denzler.
Pattern Recognition and Image Analysis. Advances in Mathematical Theory and Applications (PRIA). 24 (2): pages 243-255. 2014.
[bibtex] [pdf] [web] [abstract]

Seeing through bag-of-visual-word glasses: towards understanding quantization effects in feature extraction methods
Alexander Freytag and Johannes Rühle and Paul Bodesheim and Erik Rodner and Joachim Denzler.
ICPR Workshop on Features and Structures (FEAST): Extended Abstract + Poster Presentation. 2014. Best Poster Award
[bibtex] [pdf] [code] [presentation]

Seeing through bag-of-visual-word glasses: towards understanding quantization effects in feature extraction methods
Alexander Freytag and Johannes Rühle and Paul Bodesheim and Erik Rodner and Joachim Denzler. 2014. Technical Report TR-FSU-INF-CV-2014-01
[bibtex] [pdf] [code] [abstract]

Labeling examples that matter: Relevance-Based Active Learning with Gaussian Processes
Alexander Freytag and Erik Rodner and Paul Bodesheim and Joachim Denzler.
DAGM German Conference on Pattern Recognition (DAGM-GCPR). Pages 282-291. 2013.
[bibtex] [pdf] [web] [code] [supplementary] [abstract]

Video Segmentation by Event Detection: A Novel One-class Classification Approach
Mahesh Venkata Krishna and Paul Bodesheim and Joachim Denzler.
4th International Workshop on Image Mining. Theory and Applications (IMTA-4). 2013.
[bibtex] [pdf] [abstract]

Approximations of Gaussian Process Uncertainties for Visual Recognition Problems
Paul Bodesheim and Alexander Freytag and Erik Rodner and Joachim Denzler.
Scandinavian Conference on Image Analysis (SCIA). Pages 182-194. 2013.
[bibtex] [pdf] [web] [abstract]

Kernel Null Space Methods for Novelty Detection
Paul Bodesheim and Alexander Freytag and Erik Rodner and Michael Kemmler and Joachim Denzler.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Pages 3374-3381. 2013.
[bibtex] [pdf] [web] [code] [presentation] [abstract]

An Efficient Approximation for Gaussian Process Regression
Paul Bodesheim and Alexander Freytag and Erik Rodner and Joachim Denzler. 2013. Technical Report TR-FSU-INF-CV-2013-01
[bibtex] [pdf]

Large-Scale Gaussian Process Classification with Flexible Adaptive Histogram Kernels
Erik Rodner and Alexander Freytag and Paul Bodesheim and Joachim Denzler.
European Conference on Computer Vision (ECCV). Pages 85-98. 2012.
[bibtex] [pdf] [web] [supplementary] [abstract]

Rapid Uncertainty Computation with Gaussian Processes and Histogram Intersection Kernels
Alexander Freytag and Erik Rodner and Paul Bodesheim and Joachim Denzler.
Asian Conference on Computer Vision (ACCV). Pages 511-524. 2012. Best Paper Honorable Mention Award
[bibtex] [pdf] [web] [presentation] [abstract]

Beyond Classification - Large-scale Gaussian Process Inference and Uncertainty Prediction
Alexander Freytag and Erik Rodner and Paul Bodesheim and Joachim Denzler.
Big Data Meets Computer Vision: First International Workshop on Large Scale Visual Recognition and Retrieval (NIPS-WS). 2012. This workshop article is a short version of our ACCV 2012 paper.
[bibtex] [pdf] [abstract]

Divergence-Based One-Class Classification Using Gaussian Processes
Paul Bodesheim and Erik Rodner and Alexander Freytag and Joachim Denzler.
British Machine Vision Conference (BMVC). Pages 50.1-50.11. 2012.
[bibtex] [pdf] [web] [presentation] [abstract]

Spectral Clustering of ROIs for Object Discovery
Paul Bodesheim.
Symposium of the German Association for Pattern Recognition (DAGM). Pages 450-455. 2011.
[bibtex] [pdf] [web] [abstract]