Research Software Engineering¶
Qin Yu's work spans open-source projects in formal verification, GPU computing, computer vision, and web technologies, with particular expertise in bioimage analysis software for biological research.
Open-source¶
He is the lead developer of PanSeg, a boundary-based instance segmentation tool for volumetric microscopy images, and GoNuclear, a toolkit for nuclear segmentation in volumetric microscopy data.
He has also contributed to a range of other open-source projects:
General-purpose GPU computing¶
- CUDA.jl: Julia bindings for NVIDIA CUDA.
Formal verification¶
Max Planck Institute for Software Systems (MPI-SWS)
- CSLsound: Concurrent separation logic soundness proof.
Bioimage analysis¶
Francis Crick Institute:
- AI OnDemand: An HPC platform for segmentation workflows.
European Molecular Biology Laboratory (EMBL):
- PyTorch 3D U-Net: A U-Net framework for volumetric segmentation.
- PanSeg/PlantSeg: Boundary-based segmentation.
- GoNuclear: Nuclear segmentation.
Other communities:
- CellPose: Generalist cell segmentation.
- Ultrack: Joint tracking and segmentation.
- StarDist: Star-convex object segmentation.
- Spotiflow: Spot detection.
- magicgui: GUI framework.
Many contributions are not listed here — some due to their smaller scope, such as FiloSpot, and others because they are not publicly available.
Documentation¶
He has also written documentation for a number of tools, including: