Establishing a Decentralized and Vibrant Developer Community for ilastik

• Applicant: Anna Kreshuk, Ph.D., since 4/2018
• Subject Area: Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing ; Medical Physics, Biomedical Technology
• Term: from 2018 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 391125810

New microscopic imaging modalities open a window onto previously inaccessible phenomena in the life and material sciences. The vast amount of images produced make meaningful image analysis the new bottleneck. "ilastik" is an interactive machine learning framework for the analysis of 2D and 3D images and videos. It allows applied scientists to address the nontrivial segmentation, classification, tracking and counting tasks needed for image analysis. ilastik's core functionality is to allow training an automated analysis system by merely providing example annotations. This is especially useful for domain experts who understand what they see in the images they acquire, but have little or no experience in computer vision. Today, ilastik is the only piece of software worldwide that allows interactive training of automated image analysis on images and sequences that are much larger than the computer's main memory. It also implements state of the art algorithms for the tracking of cells and the segmentation of individual cells which are discernible only by their boundaries. As a consequence, it has been adopted by hundreds of labs worldwide that work at the cutting edge in areas ranging from the neurosciences to plant biology. However, ilastik has so far been developed by a small and homogeneous team of developers from only two institutions. Given the scope and complexity of the questions that the new imaging modalities allow to ask, a much broader base of active developers is needed for this open source project. This project will transform ilastik to make it a development platform and tool that can be sustained in the long run. To that end, four steps are necessary: First, the architecture of the current system will be overhauled to simplify it and lower the entrance barrier to new open source developers from the entire world. Second, we will provide infrastructure to facilitate cross-platform deployment. Third, a very significant effort will be made to interface ilastik with other excellent open source packages with complementary functionality. Fourth, we will reach out not only to the developers of other packages, but also educate and involve new developers in strategic decisions for the future development through workshops and hackathons. The software will continue to be open source, and the present grant will put the entire project in a shape suitable for investment by major institutions to afford new functionality and guarantee long-term survival.

• DFG Programme: Research data and software (Scientific Library Services and Information Systems)
• Co-Investigator: Professor Dr. Ullrich Köthe
• Ehemaliger Antragsteller: Professor Dr. Fred A. Hamprecht, until 4/2018