iDiv Summer School 2025: Deep learning for biodiversity and ecological research

Location


iDiv Leipzig, Puschstraße 4

Dates


25  – 29 August 2025


Recent advancements in machine learning and artificial intelligence are revolutionizing research in biodiversity and ecology. This 5-day Summer School 2025 will provide participants with a hands-on experience in applying latest deep learning and computer vision techniques to ecological and biodiversity studies.

The participants will be introduced to the theoretical foundations of deep learning and its applications in ecological research. In addition, hands-on sessions will allow participants to deepen their understanding by working with real-world examples. These practical exercises will focus on using deep learning libraries, as well as building, training, and evaluating deep learning models for ecological tasks.

The program will also feature guest presentations from researchers who will share specific case studies on topics such as data fusion, automated species identification, phenology, and integrative taxonomy using deep learning methods.

The summer school is aimed at MSc students and doctoral researchers.

We look forward to receiving your applications and hope to welcoming you soon in Leipzig.

For questions about the summer school please contact the organisation team Dr Nicole Sachmerda-Schulz and Beate Horn from yDiv Graduate School via summerschool@idiv.de.

Didactic aims and elements

Participants will acquire knowledge in diverse fields.

Technical deep learning skills:

  • Ability to use deep learning frameworks (e.g., TensorFlow, Keras, PyTorch) to build and train models.
  • Understanding of core deep learning concepts like neural networks, backpropagation, and activation functions.
  • Experience in training models on ecological data (e.g., species classification, trait recognition).

Data handling and preprocessing:

  • Competence in preprocessing ecological data (e.g., labeling, cleaning, normalization, augmentation) for deep learning tasks.
  • Experience in working with structured and unstructured data, including images, sensor data, and field observations.

Model evaluation and optimization:

  • Skills in evaluating model performance using appropriate metrics (accuracy, precision, recall, etc.) and optimizing deep learning models to improve results.

Ecological applications of AI:

  • Knowledge of how deep learning can be applied to key biodiversity issues, such as monitoring invasive species, tracking phenology and  identifying species from images.

Lecturers

idiv-idiv-

Patrick Mäder is a Professor at the Technical University of Ilmenau, where he leads the Data-Intensive Systems and Visualization group. He has developed extensive expertise in machine learning, software and safety engineering, and data mining.

idiv-idiv-

Jana Wäldchen has extensive expertise at the interface between biogeosciences, ecology and computer science. She has been conducting research at the Max Planck Institute for Biogeochemistry on the topics of monitoring ecosystems and automated species identification.

idiv-idiv-

Michael Rzanny has extensive experience in automated plant species identification, the application of deep learning techniques, and the spatial and temporal analysis of large-scale citizen science datasets, addressing a variety of research questions (e.g. phenology).

idiv-idiv-

Ladisalv Hodac has extensive experience in automated phytoplankton identification and fine grained plant species identification.

 


Costs

The total cost of  the course is EUR 150 without accommodation and EUR 350 including accommodation. This covers the costs of teaching and materials, coffee breaks and lunch, two dinners and the cultural programme. The participants will be provided a single room in a two-room apartment in the city center (Apartment Central).

In addition, 2-3 fellowships are available that will cover the registration fee as well as travel costs to Leipzig. These fellowships are for students from regions that are often underrepresented in science.

Application

Application is open until 28 February 2025.

We are currently accepting applications from students from around the world!

Please note the overall prerequisites for an application:

This summer school is meant to be an introduction to the theoretical foundations of deep learning and its applications in ecological research.

We don`t require students to have a deep knowledge of programming. The only requirements are:

  • Interest in ecological research
  • Interest in programming
  • Advanced English skills
  • Motivation to learn in an intercultural working environment

Please submit your application via the iDiv application portal at https://apply.idiv.de until 28 February 2025 and prepare to include the following information:

  1. Your name, position, home institution and contact details
  2. Please upload a letter of motivation explaining why you would like to take part in the summer school and what you like to learn (max. 500 words).
  3. In case you would like to apply for a fellowship (see section costs), provide a short statement.

The maximum number of participants is 15. The gender and nationality of applicants will be taken into account to maintain diverse participation. Application from developing countries, in particular from female applicants, will be particularly welcome.

iDiv is a family-friendly workplace, and we offer various support for families. For the summer school participants, we are able to offer on-site childcare during the daily programmme. If you require this service, please let us know when you apply.