Researchers from Imperial College London, DATAIA and Inria presented their scientific views on emerging techniques and applications of AI and machine learning as part of a programme held in partnership with the Embassy of France to the UK to develop Franco-British collaboration.
The seminar was organised following French mathematician and politician Professor Cédric Villani ’s visit to Imperial earlier this year to launch a joint mathematics laboratory with France’s National Center for Scientific Research (CNRS).
Imperial’s Professor Daniel Rueckert, Head of the Department of Computing, said that ‘artificial intelligence and machine learning is transforming how we do science and business.’
He spoke about how AI and machine learning are being used in healthcare, such as in imaging. Professor Rueckert said that machine learning can be a useful tool for radiologists and may help the get better quality images, faster.
But he warned that we are still quite far away from AI being able to diagnose a patient to the same standard as a clinician.
Professor Michael Bronstein, who started at Imperial Department of Computing that morning, gave a talk on geometric deep learning - which is using existing techniques for graph based data representation.
Professor Bronstein said: “Deep learning is coming back in a very powerful way.”
He explained that geometric deep learning has applications for self-driving cars which need to identify objects in front of them, and the detection of fake news on social media.
Driving scientific discoveriesDr Alexandre Gramfort, Researcher at INRIA in the Parietal Team and member of the DATAIA Institute programme committee, talked about the impact of open source in AI.
Dr Gramfort said ‘AI is becoming a driving effort for scientific discoveries beyond statistics and computer science departments. AI is changing the way people are testing hypotheses because they can test at larger scales and can also lead to many new scientific discoveries.”
There were also presentations from Prof. Hugues Talbot, Professor at the Centre for Numerical Vision (CVN) at CentraleSupélec, Universite Paris-Saclay, who spoke about graphical models and artificial vision, and Dr Jesse Read, Assistant Professor at the Computer Science Laboratory of école Polytechnique, who gave an interesting talk on methods deep in the output space.
French connectionsImperial has strong connections with France and have published more than 8,000 joint research papers with French institutions in the last decade. Imperial has more than 680 French students, the College’s second largest international group, and 215 French staff.
Earlier this year Imperial opened a new joint laboratory set up with France’s National Center for Scientific Research (CNRS).
The pioneering venture will integrate leading researchers and students from both institutions to significantly advance collaboration in mathematics between France and Britain.
The International Joint Research Unit - Unité mixte internationale - is named UMI Abraham de Moivre, after the great French mathematician, and will be based at Imperial’s South Kensington campus in London.
Article text (excluding photos or graphics) - Imperial College London.