Description
Uncertainty can be defined as error (outlier or deviation from a true value), imprecision (resolution of a value compared to the needed resolution), subjectivity (degree of subjective influence in the data), and non-specificity (lack of distinction for objects). Uncertainty is everywhere around us and it is also present at different steps of the visualization pipeline. In our case, we define uncertainty as any variation in visualization outcome, which is produced by an ad-hoc choice or a stochastic process, at any step of the visualization pipeline. For example, this can be noise or errors in the acquisition of the data, different alternatives in the filtering of the data, or mapping and rendering choices.
Uncertainty is an interesting phenomenon: it tends to dominate over certainty and when being part of a pipeline (such as the visualization pipeline), it is important to know how uncertainty propagates and accumulates from one step to the next, until the final outcome of the pipeline.
Tasks
- Implement and assess how uncertainty accumulates and propagates through different steps of the visualization pipeline
- Apply the developed concept of uncertainty accumulation and propagation to different scenarios of use of the visualization pipeline.
- Design additional provenance solutions and integrate them in the visualization pipeline.
Requirements
- Interest and knowledge in image processing, statistics, and (medical) visualization.
- Good programming skills.
- Creativity and enthusiasm.
Environment
To be discussed (depending on the background of the student).
References
[1] https://onlinelibrary.wiley.com/doi/pdf/10.1111/cgf.14333
[2] https://diglib.eg.org/handle/10.2312/eurorv320181143
[3] https://doi.org/10.1016/j.cag.2013.10.015
[4] https://diglib.eg.org/handle/10.2312/eurovisshort20181076