DL4MicEverywhere

DL4MicEverywhere lets microscopy researchers train and use deep learning models on their own computers through graphical Jupyter notebooks. It builds on ZeroCostDL4Mic, bringing its bioimage analysis workflows to local machines and remote computing infrastructure. You don't need coding skills to use the notebook interfaces.

The notebooks cover segmentation, image reconstruction and translation, as well as registration and denoising. Researchers can train models, check their quality and apply them to new images within the same interface. Supported workflows include U-Net for 2D multilabel segmentation.

Docker packages each notebook with a versioned software environment, which helps researchers reproduce an analysis and share it with another lab. The notebooks and models can be shared, and the platform follows BioImage Model Zoo standards for model sharing. Researchers can also add their own methods through its notebook and container framework.

The software runs on macOS, Windows and Linux using Docker Desktop. Workflows can run on a laptop or workstation, on high-performance computing infrastructure, or on a cloud system with Docker. Local execution uses your own machine; choosing a remote environment moves the computation there. Optional GPU acceleration uses NVIDIA GPUs with CUDA drivers, and processing time depends on the available computing resources.

DL4MicEverywhere is open source under the MIT license. Automated build checks validate notebook compatibility across the supported operating systems.