I lead the Lab for Mathematical Methods in Computer Vision and Machine Learning at THWS. The lab maintains the following open-source toolkits for variational methods in signal and image processing:
| Repository | What it does | Languages | Install |
|---|---|---|---|
| Pottslab | Piecewise-constant Mumford-Shah / Potts segmentation | Java, MATLAB, Python | pip install pottslab |
| L1TV | L1-TV denoising of piecewise-constant signals and images | MATLAB, Rust, Python | pip install l1tv |
| CSSD | Smoothing splines for signals with discontinuities | MATLAB, Rust, Python | pip install cssd |
| MumfordShah2D | Edge-preserving 2D smoothing via Mumford-Shah | MATLAB, Rust, Python | pip install mumfordshah2d |
| CircleMedianFilter | Fast median filter for circle-valued data (phase, orientation) | MATLAB, C++, Python | pip install pycirclemedianfilter |
| DCEBE | Bolus arrival time estimation for DCE-MRI | Python, Matlab | pip install dcebe |
| Repository | What it does | Co-author |
|---|---|---|
| HOMS_SignalProcessing | Higher-order Mumford-Shah models for joint smoothing and partitioning of 1D signals and time series | Lukas Kiefer |
| PALMS_ImagePartitioning | Piecewise affine-linear Mumford-Shah for image partitioning (online demo) | Lukas Kiefer |
| pcw-regrs | Degrees-of-freedom penalised piecewise regression | Stefan Volz |




