https://www.epo.org/en/node/cordelia-schmid?mtm_campaign=EIA2024&mtm_keyword=pressrelease&mtm_medium=press

​​Cordelia Schmid​

​​Teaching computers to “see” through machine learning​

Category
Research
Technical field
Audio-visual technology
Organisation
​​Inria​
​​Research led by computer scientist Cordelia Schmid has enabled AI to "see" and interpret complex visual data in real-time. Her work is a step towards a future in which interactive robots and self-driving vehicles provide solutions for pressing problems.​

Winner of the European Inventor Award 2024

The development of technologies enabling AI to ‘see’ and interpret complex visual data opens up new possibilities for human-computer interaction across a variety of commercial and research sectors. Cordelia Schmid’s work is significant for guiding artificial intelligence (AI) beyond basic object recognition to greater sophistication, such as understanding the context of visual scenes, interpreting human actions, and even predicting future events within a video. With the help of machine learning, which integrates vast visual, audio and text datasets, AI learns to process visual information in a similar way to humans. This enables it to recognise and categorise objects in images and video more accurately and intelligently than ever before.

A key feature of Schmid’s work is its efficiency. Visual data is processed and analysed in real-time, which is essential for applications requiring an immediate response, such as self-driving vehicles and interactive robotics. To practically implement Schmid’s research, systems must be created that can process and analyse multiple data types simultaneously, leading to faster responses. As Schmid has implemented self-improving machine learning models that consistently incorporate new data, the AI has the added capability to continuously enhance its accuracy and effectiveness.

Schmid’s research could guide AI to interact with humans with more nuance and sophistication than ever. In healthcare, for example, robots may be able to provide assistance to the elderly or detect falls.

A natural problem-solver

Cordelia Schmid was inspired by her scientific family to pursue mathematics before dedicating her career to computer vision. Her academic and professional life has been marked by pioneering contributions in the field, leading to the development of advanced image and video analysis algorithms. Schmid’s expertise has allowed her to bridge the gap between theoretical research and practical application. As the Research Director at Inria, the French national research institute for digital science and technology, she leads innovative projects. She also works part-time at Google.

Schmid is also a leading figure in the field and shapes the future of AI as an editor and chair at major journals and conferences. She applies her passion for problem-solving to the development of AI, “What motivates problem-solving for me is being presented with a difficult problem and an interesting question which you want to solve...you have to think, what are the next steps? You have to explore other solutions and attack the problem differently."

Patent numbers: