DiSSCo x WiNoDa Hackathon 2026: on machine annotation services

Mittwoch 14.10.2026, 09:00 → Donnerstag 15.10.2026, 14:00 Europe/Berlin

DiSSCo x WiNoDa Hackathon 2026: on machine annotation services (MAS)

When: 14 - 15 October (Hackathon); 14 September (Kick-Off)
Where: online
Credit: no credits
Charges: no charges
Language: English

Organisation:

The hackathon is jointly organised by the German Federation for Biological Data (GFBio) and the Naturalis Biodiversity Center.

GFBio is a leading initiative for biological data management operating across Germany and serves as a core partner of the WiNoDa Knowledge Lab (Data Science in Natural History Collections).

Naturalis Biodiversity Center, located in Leiden, The Netherlands, is a world-class research institute and natural history museum dedicated to documenting global biodiversity. Naturalis leads the development of the core data infrastructure for DiSSCo (Distributed System of Scientific Collections), which creates digital proxies of specimens in natural history collections known as Digital Specimens.

Target Group:

People who work with object-based scientific collection data and are interested in developing and connecting Machine Annotation Services (MAS) directly to the DiSSCo infrastructure.

Description:

Working in small teams, you will learn how to develop and connect Machine Annotation Services (MAS) directly to the DiSSCo infrastructure. These automated services act on Digital Specimens to provide critical data quality assessments, error corrections, or links to derived scientific data (such as gene sequences). You can work with specimen data provided by the platform or bring your own datasets (which must be integrated into the infrastructure prior to the event).

Our experts will be on hand to provide both technical and conceptual support.

For this Hackathon we specifically encourage MAS applications that do ground work for other MAS services by annotating image quality and improving AI readiness. Potential MAS applications include, but are not limited to:

 

1. ROI Segmentation: Automatically identifying and isolating areas in an image containing labels, stamps, organisms, mounting materials, rulers, or color bars.

2. AI-Readiness & Image Quality Auditing: Evaluating image clarity (blur detection), lighting uniformity, and the orientation or display state of the organism.

3. Specimen Condition Analysis: Detecting physical damage, fragmentation, or signs of pest infestation on physical objects.

4. Technical Metadata Enrichment: Extracting and validating extended metadata, such as color spaces, compression levels, resolution, and ruler to pixel calculation.

5. Historical & Contextual Analysis: Identifying collectors and curators based on handwriting recognition or identifying languages on labels.

 

The data will be processed in accordance to the Privacy Policy of the German Federation for Biological Data (GFBio). The detailed version can be found here: https://www.gfbio.org/privacy-policy.

Please feel free to contact winoda@gfbio.org if you have specific inquiries.