The mission of the Leibniz Centre for Agricultural Landscape Research (ZALF) as a nationally and internationally active research institute is to deliver solutions for an ecologically, economically and socially sustainable agriculture – together with society. ZALF is a member of the Leibniz Association and is located in Müncheberg (approx. 35 minutes by regional train from Berlin-Lichtenberg).
It also maintains a research station with further locations in Dedelow and Paulinenaue.With the frame of the BMBF-funded project, “AI and Citizen Science supported monitoring of certified biodiversity projects (German: KI und Citizen Science gestütztes Monitoring von zertifizierten Biodiversitätsprojekten) (acronym: KICS-Zert)”, the working group “Artificial Intelligence” looks for a deep learning (DL) scientist. Sustainable conservation and enhancement of biodiversity requires solutions that are scalable and consider the responsibility of economic actors and other stakeholders.
A system is needed to certify local measures for biodiversity conservation. Such a system is currently being implemented in Germany (e.g., AgoraNatura), but to support it, effective monitoring concepts are needed, which are still a major challenge for technical reasons. The use of artificial intelligence (AI; KI in German) has great potential here, but many current AI techniques, especially DL, require large amounts of biodiversity monitoring data in advance. The goal of this project is therefore to develop an AI-based biodiversity monitoring tool that can scale up biodiversity assessments for certification, and then to integrate this tool with the real-world AgoraNatura platform.
The AI tool relies on “citizen science” approaches for this purpose. To achieve this goal, in Phase 1 (1 year) we will review existing DL algorithms that meet this requirement and test the implementation of a subset of them with an existing dataset. In parallel, specific requirements for AI-assisted monitoring of conservation projects will be analyzed to link them to certification standards and match them with economic demands.
In Phase 2 (3 years if successful), the AI tool integrated into the certification system can promote a new certification model, enabling large-scale monitoring of biodiversity conservation measures environment and linking local landowners, farmers, businesses, environmentalists and citizens.
To make Phase 2 happen, we need to write a full research proposal together and pass the evaluation.Subject to funding, we are offering a full-time, 1-year (with possibility of extension) postdoctoral researcher position starting from 01.11.2023 at our location in Müncheberg as: Postdoc researcher for deep learning image analysis (f/m/d) 84-2023
Your tasks:
- facilitating interactions among collaborators for know-how transfer (monthly meeting)
- taking a leading role in full proposal writing (deadline: 30.04.2024) – upon success, a 3 year extension will be possible
- writing a review article that summarizes an overview of existing DL methods useful for the project
- applying DL algorithms which do not require any annotated images (e.g. Segment Anything Model, Semantic Clustering by Adopting Nearest Neighbors, and Self-supervised Transformer with Energy-based Graph Optimization) to analyze the LUCAS dataset.
Your qualifications:
- academic background in the field of Computer Science, Informatics, Mathematics, Environmental Science, Agriculture or related fields
- strong programming expertise in DL applications with computer vision in Python
- ability to work independently as an established scientist (proven by completed doctorate/PhD or equivalent publication successes)
- willingness to answer: What is the current best practice of DL for analyzing images without having a large number of annotations? How can computer vision be used for biodiversity conservation?
- (ideal but not mandatory) experience in using pytorch and one of the above-mentioned techniques or equivalent
- (ideal but not mandatory) fluent in German
What we offer:
- an interdisciplinary working environment that encourages independence and self-reliance
- classification according to the collective agreement of the federal states (TV-L) up to 13 (including special annual payment)
- access to modern workspace with modern IT infrastructure and HPC
- a friendly team atmosphere to work together on AI applications for sustainability
- a collegial and open-minded working atmosphere in a dynamic research institution in Müncheberg near Berlin, Germany
- good opportunity for mobile work and flexible time (negotiable)
- company ticket
Women are particularly encouraged to apply. Applications from severely disabled persons with equal qualifications are favored. It is generally possible to work in the position on a part-time basis. Please send your application preferably online (see button online application below). For e-mail applications, create a PDF document (one PDF file, max. 5 MB; packed PDF documents, archive files like zip, rar etc.
Word documents cannot be processed and therefore cannot be considered!) with the usual documents, in particular CV, proof of qualification and certificates, stating the reference number 84-2023 until 31 August 2023 to (see button e-mail application below).If you have any questions, please do not hesitate to contact us: Prof. Dr. Masahiro Ryo, Tel. +49 (0) 33432/82-206, Email: Masahiro.Ryo@zalf.de.
For cost reasons, application documents or extensive publications can only be returned if an adequately stamped envelope is attached. If you apply, we collect and process your personal data in accordance with Articles 5 and 6 of the EU GDPR only for the processing of your application and for purposes that result from possible future employment with the ZALF. Your data will be deleted after six months.