The Intermedia team studies and designs advanced communication systems between humans and machines, exploring multi-disciplinary approaches that combine digital sciences, neuroscience, and intelligent systems engineering. We focus our work on understanding, modeling, and implementing interfaces that can accurately and reliably interpret human signals, whether they are gestural, physiological, or cognitive.
We explore a wide variety of natural interfaces, such as gesture-based interaction, handwriting and pen movement recognition, interfaces based on biomedical and biometric sensors, and brain signals obtained through electroencephalography (EEG).
Our work covers the entire processing chain, from data acquisition to signal recognition and interpretation. Specifically, we are developing robust methods for signal analysis and processing to extract relevant features from complex, noisy, and multidimensional signals. We also develop machine learning and classification approaches to predict certain neurodegenerative conditions, such as Alzheimer's disease, Parkinson's disease, and dementia more broadly, from biomedical signals. Our research has practical applications in various fields, such as interpreting EEG signals to control brain-computer interfaces (BCIs), acquiring and monitoring movements for motor rehabilitation, and automating the monitoring of neurodegenerative disease progression, including Alzheimer's and Parkinson's diseases.
The Intermedia team also studies ways to prevent age-related frailty by designing systems that can detect early signs of motor, cognitive, or physiological decline. This work contributes to the development of new assistive and connected healthcare technologies designed to support the well-being and independence of older adults.
To achieve these goals, Intermedia is developing innovative statistical learning and artificial intelligence techniques, including deep neural networks, hidden Markov models, Bayesian networks, and methods for fusing multimodal and heterogeneous data. These approaches make it possible to process or even combine multiple sources of information, such as video, physiological signals, movement, and contextual data. This improves the reliability and robustness of systems that identify and interpret this information and could contribute to decision-support systems for healthcare professionals.
The team conducts its research in close collaboration with partner laboratories and medical institutions, including AP-HP, SAMU-92, Centre Hospitalier Sud-Francilien, and Hôpital Foch, as part of national and European projects. The team also contributes to research-based education by supervising doctoral students, postdoctoral researchers, and master's students, thereby fostering innovation in engineering, cognitive science, and healthcare.
The Intermedia team leverages its recognized expertise in signal processing, pattern recognition, and machine learning, and is actively involved in designing human-centered, interactive, and non-invasive technologies for future healthcare, assistive technology, and communication systems.
