Researcher Ethical Artificial Intelligence for Unmanned Aerial Vehicles
University Of Twente
In this era of big data and artificial intelligence (AI), spatial information collected through remote sensing or earth observation is essential for promoting sustainable development. Unmanned aerial vehicles (UAVs), also known as RPAS or drones, can collect very detailed imagery and photogrammetric point clouds. Machine learning and deep learning methods (e.g. CNNs, FCNs) can then automatically process this data into spatial information. Indeed this combination of spatial detail provide by UAVs with automated data processing shows great promise to support sustainable development in Lower-Middle Income Countries (LMICs), particularly in applications such as automated building detection, slum mapping, and disaster risk management.
However, there is also an increasing societal concern regarding Ethical AI or AI for good. There are questions regarding hidden biases in algorithms and concerns on how to explain the “black-box” results so they can be trusted and better utilized for decision-making. UNESCO is drafting Recommendations on the Ethics of Artificial Intelligence and the European Commission is drafting a binding Regulation for Artificial Intelligence. These guidelines emphasize the importance of embedding fairness and explainability into algorithms. And yet, despite research in the fields of FATML and XAI, there is little knowledge on how to audit algorithms in the geospatial domain for bias or best practices for explaining the results.
To address this gap there is a vacant full-time position for a researcher in deep learning / ethical AI.
The successful applicant will integrate fairness or explainability techniques into deep learning workflows for classifying UAV imagery. The precise tasks under this theme will be discussed based on the applicant’s expertise.
This position requires a background in deep learning, photogrammetry, geographical information science (GIS), earth observation and/or computer vision and must be confirmed by an excellent track record and proven programming skills.
The research output will be closely connected to the ITC UAV Centre. This Centre centralizes UAV activities at ITC and focusses on the development of educational material, capacity building projects, and consultancy services related to drones and UAVs. It is the intention that your research delivers open source code packages to be reutilized in these projects and can be integrated into the educational material produced by the Centre.
integrate fairness or explain ability techniques into deep learning workflows for classifying UAV imagery.
establish a research line on how to integrate explainable AI (XAI) or bias auditing mechanisms into deep learning workflows for Earth Observation and UAV data in particular;
implement reliable and robust deep learning workflows;
develop and share open codes for reproducible research;
publish your results in scientific journals;
present results at appropriate meetings and conferences;
be involved in educational activities and designing educational material (optional);
contribute to national and international project development (optional).
A PhD in Photogrammetry, GIS, deep learning, or computer science
An excellent track of record in one or more of these domains
Demonstrated programming skills in deep learning with open toolboxes
An ability to collaborate in a multidisciplinary environment
Excellent communication skills
An excellent command of English
An affinity with a multi-cultural education environment, excellent work ethics, and commitment to the job
Experience in writing research project proposals an teaching experience in MSc courses or equivalent are a pre.
We offer a fulltime position in an inspiring and challenging multidisciplinary and international environment for two years. Salary and conditions will be in accordance with the Collective Labor Agreement (CAO-NU) of the Dutch Universities.
Gross monthly salary between € 3,270.00 and € 4,474.00 depending on experience and qualifications (job profile Researcher 4);
A holiday allowance of 8% of the gross annual salary and a year-end bonus of 8.3%;
A solid pension scheme;
A total of 41 holiday days in case of full-time employment;
Professional and personal development programs;
Costs for moving to Enschede may be reimbursed.
Information and application
For more information about this position, you can contact Dr. Caroline Gevaert (email@example.com). You are also invited to visit our homepage.
Please submit your application before January 9, 2022 (choose “apply now” below). Your application should include:
A cover letter (maximum 2 pages), emphasizing your specific interest in the position and outlining your relevant skills and experience
A full Curriculum Vitae
Online job interviews will be scheduled on Monday January 24, 2022.