In October I visited the Big Data Value Forum (short EBDVF) which took place in Valencia this year. The EBDVF is a gathering of industry professionals, business developers, researchers, and policy-makers. The event, themed “Data and AI in Action: Sustainable Impact and Future Realities” aims to advance policy actions, and foster industrial and research activities in the realms of Data and AI. The forum is known for attracting various stakeholders connected to EU-funded projects.
On the forum I represented the project SYNTRA and its partners which is part of the EUHubs4Data program. SYNTRA aims to provide an alternative data source for training, testing, and benchmarking of AI traffic cameras. The presentation also marked the projects mid-term review, showcasing an MVP for the data pipeline created during the experiment.
SYNTRA project presentation for mid-term review
In collaboration with partners like NovelSense UG, Fraunhofer IOSB, and SDIL from Germany, Precipio from Slovenia, and the Know Center from Austria, SYNTRA has been able to harness expertise from different aspects of the AI value chain and value creation. Each entity contributes its specific knowledge, ensuring comprehensive development and implementation.
The project’s next steps involve benchmarking the synthetic data against real training and testing data, specifically data gathered by the Abakus Traffic Camera system. To enhance SYNTRA’s performance, NovelSense has plans to invest in several areas. This includes refining the surfaces of objects to make them more realistic, utilizing Generative Adversarial Networks (GANs) for creating more authentic background images, and expanding the variety of objects.
Beyond the presentation, the EBDVF served as a platform for networking, where meaningful interactions with significant players in the EU funding landscape took place.