Intelligent Vision Systems Group
"We aim at advancing the state of the art in AI, Deep Learning and Machine Learning research, while at the same time developing tailored solutions to real-world challenging problems which help to advance technology and benefit humanity."
Fields of expertise
- Computer Vision
- Machine Learning Systems (MLOps)
- Trustworthy and certifiable AI
We conduct research primarily in the domain of computer vision using 2-,3- or 4-D image or video data as input and performing classification, object detection or other visual tasks, for which we develop state of the art deep neural network architectures. We are particularly interested in recent developments including vision transformers and gauge equivariant neural networks. Domains of applications include, but are not limited to, industrial quality control, medical imaging and diagnosis (computed tomography), as well as earth (satellites) and sky (radio-astronomy) observation data. We are also interested in hybrid approaches to AI as well as geometric deep learning. Our second main area of interest concerns MLOps, which describes best practices for building complete, production-ready and scalable Machine Learning systems. Finally, we are interested in methods to create safe, trustworthy and certifiable AI systems, which comply with current and future legislation.
Services
- Insight: keynotes, trainings
- AI consultancy: workshops, expert support, advise, technology assessment
- Research and development: small to large-scale research projects, third party-funded research, student projects, commercially applicable prototypes
Team
Head of Research Group
Projects
-
certAInty – A Certification Scheme for AI systems
Certification of AI Systems by an accredited body increases trust, accelerates adoption and enables their use for safety-critical applications. We develop a Certification Scheme comprising specific requirements, criteria, measures, and technical methods for assessing Machine Learning enabled Systems. ...
-
OSR4H – Open Set Recognition for Hematology
Development of a Proof of Concept for visual Open Set Recognition (OSR) algorithms applied to a Hematology task, the classification of white blood cells.
-
AC3T – AI powered CBCT for improved Combination Cancer Therapy
The project enables a novel, combined, adaptive cancer therapy combining tumor treating field and radiation therapy due to significantly improved static (3D) and time-resolved (4D) low dose Cone Beam Computer Tomography images based on artificial intelligence image reconstruction algorithms. ...
-
Square Kilometre Array: Mock-observations via generative deep learning
In this project we use generative deep learning methods (GANs and VAEs) to produce realistic astronomical mock-observations of numerically simulated astrophysical objects, as they will be obseved by the Square Kilometre Array Telescope (SKA). This project contributes to the Swiss-wide activities for SKA of the SKACH ...
-
DIR3CT: Deep Image Reconstruction through X-Ray Projection-based 3D Learning of Computed Tomography Volumes
Project DIR3CT aims at improving the image quality of CBCT images by deep learning (DL) the 3D reconstruction from X-ray images end-to-end. This enables a novel CBCT product to be used during radiation therapy and will allow the use of these images for adaptive treatment.
Publications
-
Weng, Joanna; Reif, Monika; Chavarriaga, Ricardo; Schilling, Frank-Peter,
2023.
certAInty : a certification scheme for AI systems (Innosuisse project) [poster].
In:
Datalab Symposium, Winterthur, Schweiz, 11. Januar 2023.
ZHAW Zürcher Hochschule für Angewandte Wissenschaften.
Available from: https://doi.org/10.21256/zhaw-27261
-
Denzel, Philipp Benedikt; Gavagnin, Elena; Schilling, Frank-Peter,
2023.
Deep learning the SKA : the Square Kilometer Array project [poster].
In:
Datalab Symposium, Winterthur, Schweiz, 11. Januar 2023.
ZHAW Zürcher Hochschule für Angewandte Wissenschaften.
Available from: https://doi.org/10.21256/zhaw-27219
-
Herzig, Ivo; Paysan, Pascal; Scheib, Stefan; Züst, Alexander; Schilling, Frank-Peter; Montoya, Javier; Amirian, Mohammadreza; Stadelmann, Thilo; Eggenberger Hotz, Peter; Füchslin, Rudolf Marcel; Lichtensteiger, Lukas,
2022.
Deep learning-based simultaneous multi-phase deformable image registration of sparse 4D-CBCT [poster].
In:
AAPM Annual Meeting, Washington, DC, USA, 10-14 July 2022.
American Association of Physicists in Medicine.
pp. e325-e326.
Available from: https://doi.org/10.1002/mp.15769
-
Stadelmann, Thilo; Schilling, Frank-Peter, eds.,
2022.
Advances in deep neural networks for visual pattern recognition.
Basel:
MDPI.
Journal of Imaging ; 8.
Available from: https://www.mdpi.com/journal/jimaging/special_issues/deep_neural_network
-
Schilling, Frank-Peter; Flumini, Dandolo; Füchslin, Rudolf Marcel; Gavagnin, Elena; Geller, Armando; Quarteroni, Silvia; Stadelmann, Thilo,
2022.
Archives of Data Science, Series A.
8(2).
Available from: https://doi.org/10.5445/IR/1000146422