Nicholas Sofroniew: "napari: a Python multi-dimensional image visualization and analysis..." смотреть онлайн

Nicholas Sofroniew (Chan Zuckerberg Initiative) presenst "napari: a Python multi-dimensional image visualization and analysis platform for the research community."

Advances in imaging technologies have made it routine for biologists to generate large volume and high dimensional data that are challenging to visualize and analyze. Simultaneously, computational tools and methods for addressing these challenges have grown dramatically, including the expanded machine learning and large data handling libraries in the Scientific Python ecosystem. However, the limited large multi-dimensional image visualization options native in Python hampers researchers’ ability to fully leverage these tools effectively.

To address this gap, we are developing napari: a fast, interactive, multi-dimensional image viewer, with a vibrant plugin ecosystem that expands its capability to tackle various domain-specific visualization and analysis needs (https://napari.org/). It is built on Qt (for the GUI), vispy (for performant GPU-based rendering), and the scientific Python stack (numpy, scipy, and scikit-image).

napari is an open source project on GitHub to facilitate transparency, reuse, and extensibility (https://github.com/napari/napari). At its core, it provides critical viewer features out-of-the-box, such as support for large multi-dimensional data; provided “layers” to simultaneously visualize images, models, and analysis results; and easy manual, interactive annotation in 3D. In addition, we are aiming for flexible, developer/user-friendly plugin architecture to facilitate technology dissemination. We're also building the napari-hub (https://www.napari-hub.org/) to make it easier to discover, install, and share napari plugins. Combined napari and the napari hub thereby aim to provide biologists easy access to advanced image analysis workflows through a performant image viewer.

The Optical Interest Group (OIG) seminar series is coordinated by Ulrike Boehm and Chen Wang.