Research in Artificial Intelligence

schema
A joint private-academic institute for developing AI for the human beings and the environment


Grenoble is one of the most active places in France and Europe in core AI areas - machine learning, perception, speech communication, computer vision and embedded AI. Indeed, MIAI contributors have published, in the last three years, over 30 papers at ICML, KDD and NIPS, over 60 papers in major computer vision conferences as CVPR, ECCV and ICCV and have ranked first in several international challenges (as SemEval or ImagNet). Three European AI Fellows further contribute to research in symbolic AI. In addition, the Grenoble area includes major academic actors in hardware design, HW/SW acceleration and distributed computing.

In addition, Grenoble is at the forefront in medicine, earth and climate research. In computer assisted medical interventions, several devices developed in Grenoble have led to surgical premieres (e.g. the navigational surgery for prostate biopsies commercially developed by Koelis, a spin-off of Univ. Grenoble Alpes, that has been used by over 250 000 patients in more than 30 countries). By using massive processing of seismic and geodetic data, researchers in Grenoble have revealed that the solid earth is evolving at all time scales, an observation that is challenging the traditional views of its dynamics. By harnessing massive large-scale species distribution data and high resolution climatic and remote sensing data, researchers in Grenoble have revealed how climate and land use change triggers unexpected responses of biodiversity. MIAI will pursue and extend these research lines in several directions.

AT THE CORE OF MIAI

Future AI has to get off the cloud, in order to meet its users, and overcome problems related to communication overload and data privacy. Hardware architectures for AI (i.e. Neuro Processing Unit - NPU) is a key topic to address new applications, embedded in low power, low latency apparatus (cars, healthcare wearable devices or event smart sensors). At the same time, a new IT paradigm, mixing Edge/Fog/Cloud computing and IoT, requires advanced resource management. Distributed Intelligence is an emerging topic, which will allow optimizing distributed applications, including distributed learning. The following chairs address these issues related to embedded and distributed AI and hardware architecture for AI.
 


Automatic decision systems are currently deployed at large scale. They are already affecting the life of citizens, and their impact is expected to grow. Often based on complex data-driven machine learning models, these systems raise many scientific challenges regarding safety, robustness, privacy, fairness, and data efficiency when massively annotated data are not available. The Grenoble ecosystem possesses many assets to solve these challenges by combining the perspectives of various scientific fields, both on the academic and industrial sides. There is indeed a long tradition of research in optimisation, statistics, symbolic AI in Grenoble, as well as a more recent one in machine learning. Our strategy in MIAI consists in fostering interactions between these different disciplines, in order to make fundamental contributions to machine learning and reasoning through the chairs below
 
A major objective of artificial intelligence is to enhance the abilities of humans to interact with their environment. This involves the resolution of various problems, including perceiving, analysing and learning the informational structure of this environment, and acting on it in an adequate and efficient way. Importantly, the human environment is also composed of other humans, which raises specific questions about the automatic analysis of human behavior and the design of efficient systems for enhanced interaction between humans. The Grenoble teams have a long-standing competence on human-machine and human-human interaction, with an increasing use of machine learning techniques and maintaining at the same time ancient and strong links between computer science and cognitive psychology. The chairs bellow, addresses separately the questions of visual analysis of the external world, interaction with humans and objects in the sensory-motor framework associated to robotics, and communicating with humans by speech and language. The following chairs address these issues related to perception and interaction.
 
Artificial Intelligence offers great opportunities for elaborating innovative solutions to improve people’s life and their social environment. The integration of AI into society affects most areas of private and social lives, at the collective and individual levels. In response, individuals, groups and institutions implement regulatory processes to address the real or imagined risks resulting from AI. To avoid both disaster scenarios and the dangers of wilful blindness, social and computer scientists should join forces to implement research on the actual societal impact of AI. Moreover, reasoned regulation of AI requires not only an understanding of algorithms and technologies, but also the study of the social value and meaning that users attribute to them and the understanding of the society where they are deployed. Grenoble is particularly well prepared to meet this challenge. For several years, the Univ. Grenoble Alpes IDEX program has been promoting the development of humanities and social sciences, fostering interdisciplinary work in the digital field. Our objective within MIAI is to build on these foundations and change scale by implementing chairs dealing with the issues of: Integration of AI into society and regulation of AI by society. The following chairs address these issues related to AI and society.
 

APPLICATION DOMAINS

Health
The health and social protection systems in France and in Europe involve numerous actors: hospitals, social services, companies, universities and research and regulation centers. Data is collected from multiple sources, and while personalised patient information can revolutionize diagnosis and treatment, the real problem is capturing, storing, and understanding it. With AI, health becomes smart health, making true the promises of the P4 (Predictive, Personalized, Preemptive, Participative) vision of medicine. The CHU of Grenoble possesses both the data acquisition tools (a data lake that will be part of the French hub for AI in Health, clinical metabolomics platform) and the environment to develop and test AI-based devices (it hosts one of the 8 French Centers for Clinical Investigation and Technological Innovation). Furthermore, Grenoble is a major place of AI technologies for health, thanks to its interdisciplinary skills and its long-standing cooperation between physicians, mathematicians, computer scientists, chemists and industrial actors. Our strategy within MIAI is to enlarge these perspectives through a continuous effort on data collection and the development of new AI-based tools for omics, improved health trajectories, improved computerized assistants and augmented patient empowerment. Some of these issues are addressed in the chairs below:
 

Environment & Energy
The 2018 World Economic Forum highlighted several challenges faced by Earth systems: climate change, biodiversity and conservation, healthy oceans, water security, weather and disaster resilience. Taken together, these issues raise an urgent global challenge and require developments of AI methodologies in an interdisciplinary framework. The environment chairs of MIAI contributes to addressing all these challenges. It deals notably with the monitoring of species interactions which are impacted by climate change, the mitigation of natural disasters, the prediction of natural hazards, the forecasting of the ocean circulation and the monitoring of large hydraulic structures. Moreover, decarbonising the energy is a key issue to preserve the environment, a problem we address through the development of new technologies for smart grids.
 

Industry 4.0
Even if industry is sensor/data intensive and decision-making is at the heart of major industrial processes, the application of artificial intelligence is still nascent in manufacturing. Yet, the potentials of AI in industry are widely recognized (up to 20% cost reduction). Grenoble has a long history of research and education on industrial engineering, materials and processes and participates to the EIT Manufacturing6 on these topics. The originality of the Grenoble approach lies in the consideration that human beings are still the major industrial actors, a paradigm sometimes referred to as Industry 4.H. The aim of the chairs of this topic is to integrate AI tools in manufacturing processes to enhance product and process quality, meet the high personalization of the customers’ demand and support the promising industrial strategies, such as servitization and circular economy. A collaboration agreement with Fraunhofer (IPA and IML) has been signed to this end.