Daniel Jönsson

Associate Professor, Head of Unit

Working in the intersection between visualization and machine learning.

Making complex information understandable

Deciphering complex datasets is no easy task. Raw numbers and abstract statistics often fail to capture the essence of data, hindering human comprehension. My work aims to bridge this gap between humans and data through the power of visualization and machine learning. By combining these two fields, I try to find ways of unraveling hidden patterns, extracting valuable insights, and enabling data-driven decision-making.

The heart of my work lies in developing methods for transforming complex, high-dimensional data into intuitive visual representations. In essence, I develop visualizations that breathe life into the data and allow individuals to grasp its intricacies effortlessly. Brain imaging. Visualization of time-varying brain imaging data [D. Jönsson, IEEE TVCG, 2016], which received an honorable mention for best paper at the prestigious IEEE VIS conference. I often work in cross-disciplinary teams together with experts in fields such as neuro science, radiology, or artificial intelligence (AI), to research novel visual representations and interaction techniques. I am one of the core developers of the free rapid visualization prototyping software Inviwo [D. Jönsson, IEEE TVCG, 2019], which is used by several research groups and companies.

Daniel Jönsson. Daniel Jönsson at a workshop about education at MIT 2018. Photo credit Thor Balkhed My expertise extends beyond visualization alone. My research involves both methods for interpreting machine learning algorithms as well as making use of them to process and analyze vast amounts of data. By seamlessly integrating visualization and machine learning, I provide a comprehensive toolkit that empowers humans to extract knowledge and gain insights from large complex data.

Publications

2024

Gabriel Eilertsen, Daniel Jönsson, Jonas Unger, Anders Ynnerman (2024) Model-invariant Weight Distribution Descriptors for Visual Exploration of Neural Networks en Masse EuroVis 2024 - Short Papers (Conference paper)

Continue to DOI

2023

Alex Bäuerle, Christian van Onzenoodt, Daniel Jönsson, Timo Ropinski (2023) Semantic Hierarchical Exploration of Large Image Datasets

Christian Hansen, James Procter, Renata G Raidou, Daniel Jönsson, Thomas Höllt (Editorship) (2023) Eurographics Workshop on Visual Computing for Biology and Medicine: Frontmatter

Alex Knutsson, Jakob Unnebäck, Daniel Jönsson, Gabriel Eilertsen (2023) CDF-Based Importance Sampling and Visualization for Neural Network Training Eurographics Workshop on Visual Computing for Biology and Medicine (Conference paper)

Continue to DOI

2022

Alex Bäuerle, Daniel Jönsson, Timo Ropinski (2022) Neural Activation Patterns (NAPs): Visual Explainability of Learned Concepts

Research

Portrait of Anders Ynnerman presenting in the dome theatre.

Media and Information Technology

The research at MIT is conducted in groups that focuses on a separate aspect of visual information technology, but not isolated from each other. The collaborative atmosphere involves both basic and application oriented research.

Scientific Visualization

We develop novel analysis, visualization and exploration techniques for a better understanding and communication of large data.

Two scientists are sitting infront of an MRI

Center for Medical Image Science and Visualization (CMIV)

CMIV conducts focused front-line research providing solutions to tomorrow’s clinical issues. The CMIV mission is to develop future methods and tools for image analysis and visualization for applications within health care and medical research.

Organisation

Department of Science and Technology (ITN)

We research and teach in several technical and scientific disciplines in which our main areas are organic electronics, media, communications and logistics.