Doctoral Candidate in Computer Vision and Machine Learning
University of Zurich Department of Mathematical Modeling and Machine Learning
Switzerland
Department of Mathematical Modeling and Machine Learning (DM3L)
Doctoral Candidate in computer vision and machine learning for developing novel deep learning methods for satellite-based tracking of global CO2 and NOX emissions of point sources 80 %
Start of employment 1st of April 2026, temporaryThe EcoVision Lab in the Department of Mathematical Modeling and Machine Learning (DM3L) at University of Zurich is seeking applications for a Doctoral Candidate in computer vision and machine learning for developing novel deep learning methods for satellite-based tracking of global CO2 and NOX emissions of point sources. We offer an exciting and stimulating environment to study and work in. The University of Zurich has several internationally recognized research groups dedicated to data science, machine learning and remote sensing. We also collaborate with several other institutions and companies in the fields of computer vision, machine learning and earth observation, in Switzerland and abroad. The EcoVision Lab is member of UZH.ai, the ETH AI Center, the UZH Digital Society Initiative, the UN-ETH partnership, and the ETH for Development Center (ETH4D).
Your responsibilities
The successful candidate will work on a project in the EcoVision Lab in cooperation with colleagues at Empa and Wageningen University on developing novel deep learning methods for satellite-based tracking of global CO2 and NOX emissions of point sources (STEPS).
The next generation of polar-orbiting CO2 satellites will provide images of CO2 and NO2 emission plumes from point sources with unprecedented accuracy, resolution and coverage. The combination of CO2 and NO2 measurements will thus enable the long-term monitoring of the emissions from large point sources across the globe, which will be critical for tracking progress in reducing air pollution and achieving net-zero emissions under the Paris Agreement. The vast number of images acquired by the next generation of satellites and the large number of observable sources requires automated emission quantification methods based on deep learning.
STEPS will advance deep learning models to quantify CO2 and NOX point source emissions from CO2 and NO2 imaging satellites. The project will generate and publicly release an unprecedented library of highly realistic, globally representative satellite images based on high-resolution chemical transport simulations. The deep learning models will then be adapted to real satellite imagery to minimize the domain gap. Once developed and tested, the novel methods will be applied to the next generation of polar and geostationary satellites to monitor the CO2 and NOX emissions from power plants and industrial facilities. The STEPS project will establish an advanced framework to develop, validate and apply deep learning models for emissions quantification.
The project leaves ample room to explore various exciting technical avenues like self-supervised learning, physics-informed deep learning, uncertainty quantification, interpretability, and explainability in deep neural networks, attention-based approaches, or diffusion models, for example. Over course of the project, publications are planned at both, machine learning and computer vision conferences like CVPR, ICCV, ICLR, NeurIPS and journals like Remote Sensing of Environment, the ISPRS Journal or Nature Sustainability.
The next generation of polar-orbiting CO2 satellites will provide images of CO2 and NO2 emission plumes from point sources with unprecedented accuracy, resolution and coverage. The combination of CO2 and NO2 measurements will thus enable the long-term monitoring of the emissions from large point sources across the globe, which will be critical for tracking progress in reducing air pollution and achieving net-zero emissions under the Paris Agreement. The vast number of images acquired by the next generation of satellites and the large number of observable sources requires automated emission quantification methods based on deep learning.
STEPS will advance deep learning models to quantify CO2 and NOX point source emissions from CO2 and NO2 imaging satellites. The project will generate and publicly release an unprecedented library of highly realistic, globally representative satellite images based on high-resolution chemical transport simulations. The deep learning models will then be adapted to real satellite imagery to minimize the domain gap. Once developed and tested, the novel methods will be applied to the next generation of polar and geostationary satellites to monitor the CO2 and NOX emissions from power plants and industrial facilities. The STEPS project will establish an advanced framework to develop, validate and apply deep learning models for emissions quantification.
The project leaves ample room to explore various exciting technical avenues like self-supervised learning, physics-informed deep learning, uncertainty quantification, interpretability, and explainability in deep neural networks, attention-based approaches, or diffusion models, for example. Over course of the project, publications are planned at both, machine learning and computer vision conferences like CVPR, ICCV, ICLR, NeurIPS and journals like Remote Sensing of Environment, the ISPRS Journal or Nature Sustainability.
Your profile
We are looking for candidates with an interest in performing innovative research, strong motivation, and an interest in software development. An ideal candidate will have:
- an excellent degree (M.Sc. or equivalent) in Computer Science, Machine Learning, or a related field (e.g. Electrical Engineering, Applied Mathematics, Physics)
- strong understanding of maths and physics
- experience in programming, preferably in Python
- prior experience in machine learning, computer vision and remote sensing and strong interest to apply these skills to an interdisciplinary project
Information on your application
Please submit your complete application (motivation letter, curriculum vitae, school and university score records, contact details of at least two referees) via the link below. The deadline for applications is 30.11.2025 and the desired starting date is 01.04.2026. Selection will start immediately, so early submissions are encouraged.What we offer
Work-Life Balance
- Flexible working models (such as part-time positions, mobile working, job-sharing)
- Childcare at the kihz foundation of UZH and ETH
Learning and Development
- Wide range of continuing education courses of UZH and the Canton of Zurich
- Language Center run jointly with ETH Zurich
Food
- Food and drinks at reduced prices in the UZH cafeterias
- Lunch-Check-card with UZH contribution
Healthcare
- Special conditions on the Academic Sports Association ASVZ
- Free seasonal flu vaccinations
- Rest and relaxation at the quiet room in the university tower
Discounts
- Private traffic: Carsharing, rent a vehicle, parking space
- Digitalization: Hardware, software, mobile phone subscriptions
- Special conditions on hotel reservations
Conditions of Employment
- Policies of the UZH
- Most UZH staff are employed according to public law
International Services
- Support for people from outside Switzerland
Campuses
- Campuses Zurich City, Zurich Irchel, Oerlikon and Schlieren
- Sites Zurich West, Old Botanical Garden, Botanical Garden and Lengg
Location
Department of Mathematical Modeling and Machine Learning (DM3L)
Winterthurerstrasse 190, 8057 Zürich, SwitzerlandFurther information
Questions about the job
Prof. Jan Dirk Wegner ProfessorThank you for your message. We will get back to you shortly.
Questions about the application procedure
Nicole Trolese HR ManagerThank you for your message. We will get back to you shortly.
Working at UZH
The University of Zurich, Switzerland's largest university, offers a range of attractive positions in various subject areas and professional fields. With around 10,000 employees and currently 12 professional apprenticeship streams the University offers an inspiring working environment on cutting-edge research and top-class education. Put your talent and skills to work with us. Find out more about UZH as an employer! More
In your application, please refer to Professorpositions.com
