Join us for the AI4Life Community Event, a milestone celebration bringing together contributors, users, partners, and community members to reflect on the project’s achievements and explore future collaborations.
Location: Katajanokan Kasino, Laivastokatu 1, 00160 Helsinki, Finland
Tentative Agenda
Tuesday, May 27th - Katajanokan Kasino, Kenraalisali
(On-site)
Workshop 1: AI4Life for life scientist
| 09:15 | Registration |
| 09:30 | Session 1: Introduction to the BioImage Model Zoo: AI tools for image analysis in the life sciences |
| 10:30 | Coffee break |
| 11:00 | Session 2: AI4Life community spotlight: Connecting users with community partners |
| 12:00 | Session 3: Getting your data AI-ready at the BioImage Archive |
| 12:30 | Lunch |
Workshop 2: AI4Life for AI scientists, software developers and image analysts
| 12:30 | Lunch | ||||||||
| 13:45 | Registration | ||||||||
| 14:00 | Session 1: Creating, Uploading, Deploying and Containerizing BioImage Model Zoo Models | ||||||||
| 15:00 | Session 2: Bring BioEngine In-House: Self-Hosted AI Models for Bioimage Analysis | ||||||||
| 16:00 | Coffee break | ||||||||
| 16:30 | Session 3: Present and future of data and AI model standards
| 17:15 | Session 4: So what’s this all good for? A story about the potential of shared and FAIR resources | 17:45 | End of the Workshops | 17:45 | Registration | 18:00 | AI4Life Community Event Dinner and Networking at Katajanokan Kasino, Laivastokatu 1, 00160 Helsinki | |
Wednesday, May 28th - Katajanokan Kasino, Merisali
(Hybrid)
| 09:00 | Registration |
| 09:30 | Session 1: Welcome and Introduction to the AI4Life Project |
| 10:15 | Group picture & Coffee break |
| 10:45 | Session 2: AI4Life Success Stories |
| 12:30 | Lunch |
| 13:30 | Session 3: The AI4Life Community |
| 14:30 | Coffee Break |
| 15:00 | Session 4: AI4Life and the European Open Science and AI Landscape |
| 15:30 | Discussion, Wrap up & Outlook |
| 16:00 | Official End of the AI4Life Community Event |
Abstracts
Workshop 1: AI4Life for life scientist
Session 1: Introduction to the BioImage Model Zoo: AI tools for image analysis in the life sciences
The BioImage Model Zoo is the heart of the AI4Life project. It holds a collection of community-contributed neural networks, pre-trained to perform tasks of microscopy image analysis such as nucleus or cell segmentation. This session will introduce the Model Zoo website and demonstrate how to explore models, try them on your own data and download them for offline use. We will also show the BioImage.io Chatbot that can help with your image analysis tasks.
Session 2: AI4Life community spotlight: Connecting users with community partners
The AI4Life community is built on collaboration, bringing together diverse partners who develop and support AI-powered bioimage analysis tools. In this session, community partners such as BiaPy, SpotMAX, QuPath and DeepIcy, will introduce themselves, showcasing their contributions to the AI4Life ecosystem. Participants will gain insights into the tools and resources available to enhance their research workflows, engage with the developers behind these technologies, and explore the opportunities for collaboration.
Session 3: Getting your data AI-ready at the BioImage Archive
AI and machine learning have the potential to accelerate scientific discovery by allowing life scientists to extract meaningful information from microscopy data. To be able to develop AI models, computational scientists need access to well-annotated images. However, for image datasets to be useful to AI scientists, they need to be FAIR (Findable, Accessible, Interoperable, and Reusable). One of the aims of the BioImage Archive (BIA), EMBL-EBI’s repository for biological images, is to serve as a FAIR source of AI data.
In this workshop we will explore how the BIA can help life scientists with making image data FAIR and AI-ready, and with choosing AI-models that will perform well on their images. First, we will learn how the BIA is making annotated datasets FAIR and AI-ready. As part of this, we will navigate our AI Gallery, a collection of datasets that has been curated to be extra useful to AI-scientists. Then we will be able to test how different AI models from the BioImage Model Zoo perform on datasets from the BIA. You will learn about different metrics that can help you make decisions when choosing an AI model for your images.
Workshop 2: AI4Life for AI scientists, software developers and image analysts
Session 1: Creating, Uploading, Deploying and Containerizing BioImage Model Zoo Models
The BioImage Model Zoo (BMZ) is a community-driven platform designed to make deep learning models for bioimage analysis accessible, interoperable, and reusable. At its core, the BMZ employs standardized specifications to describe models, datasets, and applications, ensuring cross-compatibility across various bioimage analysis tools (Community Partners). These specifications encapsulate essential metadata enabling seamless integration into diverse image analysis workflows.
This workshop will provide participants with practical knowledge of the BMZ ecosystem in Python. Attendees will learn how to create, validate, and share models using dedicated packages that facilitate interaction with the BMZ specifications. The session will also cover deploying and integrating these models into larger pipelines using the BMZ containerization system and its implementation via DL4MicEverywhere. Through hands-on exercises, participants will gain the skills needed to contribute to the BMZ and leverage its resources for advanced bioimage analysis.
Session 2: Bring BioEngine In-House: Self-Hosted AI Models for Bioimage Analysis
BioEngine is a user-friendly, cloud-based platform that provides a seamless way to serve and run AI models from the BioImage Model Zoo on your own data. By integrating with popular software tools like Fiji, Icy, and napari, BioEngine greatly simplifies complex software setups and automates GPU resource management—ideal for facilities, Euro-BioImaging Nodes, industry partners, and developers alike.
In this workshop, we will walk participants through the steps to deploy and configure BioEngine on both local workstations and cluster environments, ensuring fast, scalable AI inference on a wide range of bioimage data. Attendees will learn best practices for installation, discover how to integrate BioEngine into existing hardware and software infrastructures, and gain insights into extending its capabilities to meet various research and business needs.
Session 4: So what’s this all good for? A story about the potential of shared and FAIR resources
In this session, we would like to zoom out a bit and show what AI4Life and Euro-BioImaging, the European infrastructure for biological and biomedical imaging, offer today, and why shared infrastructures will unlock a brighter future in AI and with this also have a positive impact also for pushing areas from basic science to industrial applications.