IJmondCAM aims to train an Artificial Intelligence model to spot toxic clouds, and we need volunteers to help label smoke emissions. So far, [...] videos are fully labeled. We would greatly appreciate your participation in this community-powered AI research!

Our goal is to invite citizens in creating data-driven evidence of pollution events (using the smoke emissions found by the AI model). We hope that this evidence can facilitate air quality advocacy in the IJmond region in the Netherlands.

The system supports Android 7+, iOS 11+, and modern desktop browsers. More information about this project can be found in the highlighted FAQs. For technical details, please refer to another set of FAQs. We appreciate your support!


The original camera footage data are provided by Frisse Wind. The videos are then processed by the data engineering team in the AI against Toxic Clouds challenge. The code for this web-based tool is open-sourced on GitHub.

Video autoplay is disabled. Please enable it.