Review of the official presentation event of Confiance.ai
On 1 July, the members of the Confiance.ai collective, made up of 13 major French industrial and academic partners, met at the Valeo Mobility Tech Center in Créteil. On the programme: highlighting the industrial issues and scientific challenges at the heart of this programme.
The morning, attended by over 300 participants, was marked by several highlights:
- The opening remarks by Marc Vrecko, Director of Valeo’s Comfort and Driving Assistance Systems Business Group, Julien Chiaroni, Director of the Great Challenge “Securing, certifying and making reliable systems based on artificial intelligence”, Géraldine Leveau, Deputy Secretary General for Investment, and Michel Morvan, President of IRT SystemX.
- The opening speech, given by Cédric O, Secretary of State for Digital Transition and Electronic Communications. In particular, he highlighted the fact that “the field of trusted AI, and more globally of AI applied to industrial data, is one of the areas in which we must and can be one of the world’s champions.”
- The members involved in Confiance.ai collective then shared their keywords on video:
- Presentations, by members of the collective, on industrial issues and scientific challenges of the programme.
- A focus on the manifesto for AI at the service of industry signed in July 2019 was then made by Marko Erman, Scientific Director of Thales. Its objective: to make AI a source of growth and employment at the service of industry, particularly in terms of massive data, which is critical for competitiveness.
- A round table dedicated to the European and international synergies of the Confiance.ai programme presented the collaborations undertaken by the group, in particular with Germany and Quebec.
- The event was concluded by Jean-Baptiste Djebarri, Minister Delegate to the Minister for Ecological Transition, in charge of Transport. During his speech, the Minister stressed the importance of trusting AI to support the digital transformation of the transport sector.
Watch the whole event in replay:
Innovating for IA
The Confiance.ai programme makes it possible to create unique synergies between the worlds of research and industry, with the aim of rethinking the entire design chain of a trusted AI. During the event, four demonstrations were featured to discover part of the R&D work carried out by members of the collective:
Building databases for trusted AI – The Confiance.ai Programme
The Confiance.ai collective presented an AI tool for annotating objects on an image, thus reducing the cost of development and improving the AI’s understanding of a scene to guarantee trust, and a tool for detecting bias in data.
SaaS solution to measure the robustness of an AI and extract explainability elements – Numalis
Numalis shared tools to ensure AI reliability by measuring robustness and explaining behaviour.
SimfiaNeo: identification of safety sequences for the autonomous vehicle – Apsys
Apsys presented its SimfiaNeo platform, which uses modelling, simulation and a formal language called Altarica to identify the use cases of an autonomous vehicle and to highlight those leading to an accident, taking into account environmental factors or other external phenomena.
3D object detection by Deep Learning for autonomous driving – Valeo
Valeo demonstrated AI perception and presented the DeepManta software (developed in collaboration with CEA) for multi-sensor perception (cameras and lidar cloud point).