Four PostDoc Positions on AI4EO for Social Good
The TUM Chair of Data Science in Earth Observation (SiPEO) develops innovative signal processing and machine learning algorithms to extract geo-information from big geospatial data, ranging from remote sensing satellite data, aerial data, LiDAR to social media data. As downstream applications, we provide large scale and highly accurate geo-information to address societal grand challenges, such as monitoring the global urbanization, climate research and supporting the sustainable development goals of the United Nations. Towards this goal, our chair hosts large scale initiatives, such as the international future lab AI4EO, the national excellence center ML4Earth and the TUM innovation network EarthCare, and thus, offers a close connection to a large AI4EO research team covering a wide range of expertise ranging from machine learning/deep learning, remote sensing and Earth observation, big data analytics, data mining, data fusion, HPC to sustainability. In addition, we offer analysis-ready data on a global scale that are either open and free, such as Sentinels, or accessible through scientific proposals or academic collaborations, such as TanDEM-X and Planet Scope.
To establish the AI4EO for Social Good working group, our lab offers four open positions for outstanding postdocs and senior scientists (initially for two years with the possibility to prolongate). Interested applicants shall share the vision of the working group. The main research tasks will be AI4EO for addressing the challenges of our time, building up on the expertise and data available at the chair. The exact topic will be defined jointly by the candidate and Prof. Xiaoxiang Zhu – the PI of the lab.
Interested applicants should have an excellent track record and a PhD in machine learning, computer science, statistics, remote sensing, mathematics or a related discipline.
Application materials comprise:
− Full set of transcripts
− Statement of purpose
− Briefly state what drives you and what are your goals in applying to the SiPEO lab
− Names for at least two references (For each reference, please include name, title, and email address. References should expect to be contacted for a reference letter.)
Please submit these documents to email@example.com by November 30th, 2022. Please kindly consider that due to the high requests, we will not be able to consider incomplete applications.
Prof. Xiaoxiang Zhu
Technical University of Munich
Chair of Data Science in Earth Observation
80333 Munich, Germany
PhD and PostDoc Positions in AI4EO
PhD and PostDoc (Wissenschaftliche/r Mitarbeiter/in)
Zhu lab develops innovative signal processing and machine learning algorithms to extract geo-information from big geospatial data, ranging from remote sensing satellite data and even social media data. As downstream applications, we provide large scale and highly accurate geo-information to address societal grand challenges, such as monitoring the global urbanization, climate research and supporting the sustainable development goals of the United Nations. Our lab offers currently several open positions for outstanding PhD Candidates, postdocs, and senior scientists. We also have open positions for outstanding research engineers.
Topics of particular interest to the group include:
- Earth Observation and Computer Vision
- Machine Learning/Deep Learning
- Unsupervised/weakly Supervised Learning
- Uncertainty Analysis, Interpretation and Reasoning of Deep Neural Networks
- Anomaly and Change Detection Methods
- Geo-information Extraction from Social Media Data
- Natural Language Processing
- Large-Scale Data Mining and Knowledge Discovery in Earth Observation
- Big Data Management
- High-performance Computing
- Statistical Learning, Modelling, Spatial and Temporal Analysis of Geographical Observations
- Geo-referencing, Digitalization, Building and Maintaining Large Relational Data Bases and Geo-databases, Publishing Geo-services (i.e. Web Map Services)
Demonstrated hands-on experience in one or more of these areas is a requirement. Postdoc applicants should have an excellent publication record and a PhD in machine learning, computer science, statistics, remote sensing, mathematics or a related discipline. Research engineer applicants should have excellent coding skills, as well as practical skills in data science and/or deep learning, and experience with scripting and running large-scale experiments.
Application materials comprise:
- Full set of transcripts
- Statement of purpose
- Briefly state what drives you and what are your goals in applying to the SiPEO lab
- Names for at least 2 reference letter writers
For each reference, please include name, title, and email address.
References should expect to be contacted for a reference letter.
Please submit these documents to firstname.lastname@example.org. Please kindly consider that due to the high requests, we will not be able to consider incomplete applications.
Science Manager (m/f/d)
The International Future Lab “AI4EO: Artificial Intelligence for Earth Observation – Reasoning, Uncertainties, Ethics and Beyond” (https://ai4eo.de/) brings renowned international organizations across the globe and a rich number of highly ranked scientists at all levels together to address three fundamental challenges in Earth observation specific cutting-edge artificial intelligence research –Reasoning, Uncertainties, and Ethics. The research carried out in the Future Lab AI4EO will not only advance Earth observation science but also make key contributions for the interpretability of AI and its ethical implications, and towards AI4EO technology transfer.
TUM serves as the host institution for the Future Lab AI4EO, represented by the Professorship “Data Science in Earth Observation” (Prof. Xiaoxiang Zhu). Consistently ranked among Europe’s leading universities, TUM has recognized the challenges emerging in the digital age and is already spearheading the advancement of Al research from fundamental stages via applied research to the study of Al’s social implications. In this context, the DLR is a strong partner. Together with LMU, Helmholtz Center Munich and Max Planck Institute for Plasma Physics, DLR and TUM established the Munich School of Data Science (MUDS) to train the next generation of “data scientists” – interdisciplinary experts in applying, adapting, and developing methods for AI and data science tailored for a broad array of research domains, including EO. In addition, DLR founded the local Helmholtz Artificial Intelligence Cooperation Unit (HAICU) – MASTr (Munich Unit @ Aeronautics, Space and Transport), which is providing AI expertise from EO, robotics, computer vision and HPC/HPDA support. DLR also started strategic cooperation with Leibniz Supercomputing Centre, e.g. through recently signed cooperation agreement “Terra Byte”, which shall enable the highly efficient and independent analysis of large amounts of data using the latest methods to understand global trends and their consequences. This further strengthens the already existing network of collaboration.
The Lab will be physically located at the new campus of TUM in Taufkirchen/Ottobrunn, where TUM currently has established its new Department of Aerospace and Geodesy as part of the Bavarian space initiative. Besides TUM, the area also hosts the University of Federal Armed Forces, Munich Aerospace, and several EO and space industry partners (e.g. Airbus, Siemens, and IABG), thus providing the ideal location for knowledge and technology transfer. In addition, two further institutions, namely Munich Data Science Institute (MDSI) and the TUM Institute on Ethics in AI, are strong partners, who will provide data science theoretical support and ethical guidance to AI4EO.
The highly visible position of the Lab with its internationality and the close connection to other Institutes requires a respective Management. You will be in close contact with our international Guest Professors and our motivated Young Fellows, as well as with colleagues from the partners form the German Aerospace Center and the TUM
Through this advertised position the Future lab is expected to be supported mainly through appropriate science management, as well as to seek new collaborations. You will also initiate, implement and manage new scientific projects in the field of artificial intelligence for earth observation.
In addition, independent, scientific research on the topic of AI for Earth observation will be supported.
- PhD in computer science, geoinformatics, data science, business administration, or comparable field of study
- professional experience in Earth observation and/or Data Science
- Background of general AI Methods, ideally applied on EO data
- very good knowledge in the area of scientific project application as well as the German research landscape
- good experience in the field of scientific communication ('Science Communication')
- experience in project management
- fluent command of the German language and very good knowledge of the English language, both written and spoken
- excellent communication and cooperation skills, ability to interact with scientists at different levels
- Fluent in spoken and written German language
- very good knowledge of spoken and written English language
- ability to work highly motivated and independently in a team
OTHER WELCOMED QUALIFICATIONS
- good graphic design skills with tools such powerpoint and photoshop is welcomed
- experience in website design
Payment will be based on the Collective Agreement for the Civil Service of the Länder (TV-L, up to E14 level). TUM strives to raise the proportion of women in its workforce and explicitly encourages applications from qualified women. Applications from disabled persons with essentially the same qualifications will be given preference.
Contact: Interested candidate please send your motivation letter, CV, and supporting materials with the key word Science Manager to email@example.com. For further questions on this position, please contact Prof. Xiaoxiang Zhu (firstname.lastname@example.org).
10 new research, data science, and management positions at Technical University of Munich in novel center for Machine Learning in Earth Observation (ML4Earth)
AI methods, and especially machine learning (ML) with deep neural networks, have replaced traditional data analysis methods in recent years. The Technical University of Munich (TUM), together with the German Aerospace Center’s Remote Sensing Technology Institute, has created the biggest European research team on AI for Earth Observation (AI4EO) over the past few years.
Starting in 2022, we will begin establishing a national ML4Earth center of excellence with high visibility. It will conduct own research at the highest international level by tackling fundamental methodical challenges in AI4EO and their application to the European mission of a Digital Twin Earth. ML research directions will include physics-aware machine learning, reasoning, uncertainty estimation, Explainable AI, Sparse Labels and Transferability, as well as Deep Learning for Complex Structures. These novel methods will be applied to practical tasks such as predicting European water storage, quantifying permafrost thawing, sea level budget, climate and earth system modeling, soil parameter mapping, and multi-sensor segmentation, together with our partners at renowned international and national institutes such as Bonn University, Alfred Wegener Institute, University of Bristol, Leipzig University, and the German Aerospace Center.
Another important goal of the project is international community building within the AI4EO domain. Aspects of this include the creation of benchmark data sets for a wide range of application scenarios, as well as offering educational resources for interested scientists and expert workshops. Our aim is the democratization of AI4EO to enable more researchers to exploit Copernicus and other EO data sources. To establish this new center, we are currently offering:
- 6 new PhD candidates/Postdocs in the described scientific topics (all filled)
- One project manager (open)
- One team assistant (50%) (open)
- One HPDA/HPC support scientist (open)
- Benchmarking and education roles (filled)
For more information, please contact us under email@example.com.
Teamassistenz (m/w/d) 50%
- Betreuung und Unterstützung der Team-Mitglieder in allen administrativen Finanz- und Personalangelegenheiten (Urlaub, Krankmeldungen, Reisekostenabrechnungen, Bestellungen)
- Kommunikation mit der Verwaltung der Technischen Universität München
- Bearbeitung von Einstellungsvorschlägen
- Durchführung von Buchungen in SAP R/3
- Unterstützung des Projektmanagements bei der Verwaltung des Budgets des Zukunftslabors und der Erstellung von Projektberichten in deutscher Sprache
- Selbständiges und verantwortungsvolles Arbeiten im Team
- Abgeschlossene kaufmännische Berufsausbildung oder vergleichbare Erfahrung mit Verwaltungsaufgaben
- Kenntnisse in SAP R/3
- Sicherer Umgang MS-Office-Anwendungen
- Sehr gute Deutsch- und Englischkenntnisse
- Schnelle Auffassungsgabe, Kommunikationsfähigkeit und selbständiges Arbeiten
Wir wünschen uns eine aufgeschlossene und teamfähige Persönlichkeit mit einem freundlichen und professionellen Umgang. Erfahrungen im universitären bzw. internationalen Wissenschaftsumfeld sind vor Vorteil. Schwerbehinderte werden bei im Wesentlichen gleicher Eignung und Qualifikation bevorzugt eingestellt. Die TUM strebt eine Erhöhung des Frauenanteils an, Bewerbungen von Frauen werden daher ausdrücklich begrüßt.
Die Stelle ist ab sofort zu besetzen und zunächst bis 31. August 2024 (Laufzeit der Projektförderung des Zukunftslabors) befristet. Die Vergütung erfolgt nach TV-L E8 (50%).
Bewerbung: Wir freuen uns auf Ihre vollständigen und aussagekräftigen Unterlagen (inkl. Zeugnisse) per email in einer PDF-Datei unter dem Stichwort "AI4EO: Teamassistenz" an firstname.lastname@example.org.
Open Positions at DLR
4 PhD or Postdoc Positions at the Remote Sensing Technology Institute at the German Aerospace Center (DLR).
The department "EO Data Science" at the Remote Sensing Institute of DLR, located in Oberpfaffenhofen near Munich, develops advanced signal processing and AI techniques for current and future Earth observation missions. It is involved in numerous third-party funded projects and a large international network.
In the context of our project work, we have 4 open scientist positions to fill. The outcome of a PhD degree for these positions is foreseen.
Position "WeMonitor" - duration 3 years, starting as soon as possible:
The tasks to be worked on are directly relevant to the timely recording of changes on the entire surface of the Earth. Socially highly relevant examples are (illegal) waste dumps, landslides, deforestation, dam failures, fire, or volcanic activity. Due to their short time scales, these events have to be distinguished from seasonal long-term changes by precise analyses of the spatial as well as temporal domain. To accomplish this efficiently, the core task is scientific research on Artificial Intelligence for Earth Observation and the development of sophisticated, relevant Deep Learning (DL) algorithms. Project management is an equally important part of the task profile.
The DL model development tasks can be further broken down into:
- Development of interactive data annotation methods for training weakly supervised DL models.
- Implementation of a transfer learning algorithm to benefit from parallels to other research fields regarding spatio-temporal models for anomaly detection.
- Development of an efficient method of multi-modal learning.
Due to the long experience of the Geoforschungszentrum Potsdam (GFZ) in the evaluation of satellite data and in the relevant scientific fields, the GFZ takes over the evaluation of the data products and the application of the models developed at DLR for active learning, model transfer, and evaluation and validation of the models. A close exchange with the GFZ is therefore indispensable.
Position "AutoCoast" - duration 3 years, starting as soon as possible:
Much of the world's coastal regions are undergoing severe changes due to either erosion or sea level rise. Due to the dense settlement of these regions, there is an acute need for action, especially to quantify the rate of change.
The AutoCoast project aims to achieve this by developing advanced, reliable AI algorithms based on high-resolution Earth observation data. Before the developed methods are applied globally, the Baltic Sea and North Sea will be used for study and validation purposes.
The project has two specific objectives:
- Classification of shoreline types based on multiple sensors onboard Earth observation satellites (e.g., sandy beach, cliff, wetland, ...) and quantification of rates of change.
- Identification of local causes of such changes using Explainable AI.
The partner Hereon will subsequently integrate the results into the marine geoportal "coastMap". The Helmholtz Center Hereon is an interdisciplinary center for coastal research. A close exchange with Hereon in the course of the project is indispensable.
Position "SURF" - duration 3 years, starting as soon as possible:
The tasks to be worked on are directly relevant to the rapid recording of flood damage after flood
events - a highly relevant and explosive field of activity for the future. The core task is scientific
research on the topic of Artificial Intelligence for Earth Observation and the development of
sophisticated, relevant Machine Learning (ML) algorithms.
The main focus of the tasks is on:
- Developing an unsupervised ML model for change detection based on time series data from Earth observation satellites. The direct goal is to support flood management.
- Development of ML algorithms for segmentation of building footprints, especially for damaged buildings. This special case presents an additional challenge. An envisioned solution approach is to exploit manifold data from various sensors on board different earth observation satellites, as well as time series of these data.
- Develop a semi-supervised model based on the unsupervised ML model from point (1). Data annotation is a particularly costly challenge for change detection models based on time series data. The goal is to detect different patterns of change in different geographic environments. It must be assumed that pre-annotated data are scarce.
The evaluation of the data products and single case applications of the ML algorithms are driven
by the Geoforschungszentrum (GFZ) Potsdam. A close exchange with the GFZ is therefore
Position "Evoland" - duration 3 years, starting as soon as possible:
The abundance of Earth observation data from radar measurements, multi- & hyper-spectral imagery, and spaceborne LiDAR surveys provides ample opportunities to sense the surface of our planet.
An increasing amount of success stories proves artificial intelligence methodologies useful for remote sensing applications. Our team "Large-Scale Data Mining in Earth Observation" (DM4EO) applies state-of-the-art weakly-supervised learning to tackle real-world challenges such as accurate land surface mapping for the European Copernicus programme (https://land.copernicus.eu/global/products/lc) under the constraint of limited amounts of labels available.
We invite motivated professionals with a Masters degree from physics, mathematics, data science, and geoinformatics to join our mission to advance scalable technologies with the aid of machine/deep learning. In particular, you would get familiar with data fusion and analytics of sensors such as DLR's EnMap satellite (hyper-spectral), ESA's Sentinel-1/2 missions (radar and multispectral), and NASA's GEDI instrument (spaceborne LiDAR). Given coding skills in Python, preferably with initial expertise in the PyTorch framework, you will perform research in novel methodologies to improve semantic segmentation for land cover monitoring such as forests and agricultural land. More specifically you will tackle challenges in spatio-temporal data fusion for change detection from a plurality of satellite sensors with the aid of self-supervised learning.
Besides your affiliation with DLR, your project offers opportunities to interact with relevant European governmental and private institutions such as the French Space Agency, Vito (BE) and GAF AG (DE). Moreover, DM4EO actively seeks collaboration with overseas, among: IBM TJ Watson Research Center, NY and the World Bank in Washington, DC. You will acquire expertise in deep learning research for computer vision in the geosciences. You will not only conduct application-relevant science, but your skill set is going to adopt industry-standard coding for future track record in either academia or industry.
Qualifications for all positions:
- Completed scientific university studies (university diploma/master) in computer science, data
- science, geoinformatics, geodesy, physics, or comparable disciplines
- Mature programming skills in Python with hands-on experience in PyTorch
- Experience in machine learning (ML) and big data handling.
- Ability to clearly dissect the essence of mathematical methods and numerical algorithms
- Experience in computer vision and remote sensing is an advantage
- Very good knowledge of the English language, both written and spoken
- Strong internal and external communication and collaboration skills
- Ability to interact with scientists at different levels
- Ability to work both independently and as part of a highly motivated team
Please send your full application (motivation letter, CV and certificates) with mentioning your favorite position to email@example.com