Search

To meet the needs of the hybrid communities, Hypermedia MAS need to address the following issues: - world-wide: Hypermedia MAS have to accommodate billions of people and autonomous agents, to scale up to the size of the Web, and to cope with the high latency inherent in world-wide systems (e.g., r

Reactive ML algorithms adapt to data generating processes, typically do not require large computational power and, moreover, can be translated into offline (as opposed to online) algorithms if needed. Introduced in the 30s in the context of clinical trials, online ML algorithms have been gaining a

Evaluating search systems requires setting up an evaluation environment: select a paradigm, metrics, a dataset, etc. The choice of an environment is rarely motivated objectively, and the impact of its variations (choosing a dataset against another, altering one, etc.) is rarely measured. Such object

In the project, we consider three types of modality, namely (1) the visual modality extracted from pixels of the images (2) the textual modality extracted from the questions in natural language, the captions and other textual contents that are “near” an image and the textual documents used to pop

Autonomous driving is set to have a profound impact throughout our entire society in the near future. However, several scientific challenges remain unanswered, such as the automated analysis of dynamic 3D point clouds from moving vehicles at a precision and speed compatible with a fully-autonomous

We combine machine learning (ML) and control theory (CT) and address problems in control from two perspectives: - What can be modelled (CT) and what needs to be learned (CT)? - Can we provide, estimate or guarantee stability (CT)? - Can we estimate the complexity of the learning task and or the a

Today's overabundance of textual data makes it difficult to find the correct answer to a user query. In this context, Question Generation (the ability to automatically generate questions from a document) is rapidly gaining traction as a key technology. This project aims to investigate the task of pa

Web search, drug design and traffic management are only a few examples of crucial applications that nowadays rely on the analysis of graphs. Involved graphs have become increasingly massive sometimes reaching trillions of edges such as the Web graph, Facebook, Internet or a human brain. Only quasi-l

The last ten years have witnessed considerable expansion into various omics data that has resulted in an explosion of publicly available heterogeneous datasets. Recent genotyping and profiling technologies enable the scientific community to investigate disease-related genomic alterations in human di

Audition is a key modality to understand and to interact with our spatial environment, and plays a major role in Augmented Reality (AR) applications. The HAIKUS project investigates the use of Artificial Intelligence (AI) for synthesising augmented acoustic scenes. Embedding computer-generated or pr

The goal of the PsotProdLEAP project is to develop new tools for video archive post-production by leveraging both recent deep learning approaches and patch-based and variational approaches. Frequent artefacts observed with deep-learning methods include loss of details, spatial and temporal discontin

With most deep learning (DL) architectures and algorithms having developed for 2D images, their adaptation to 3D data (point clouds or meshes) is less obvious, where a regular structure is not directly available. While DL CNNs have been used in some 3D contexts, e.g. face modeling or shape classific

WP1: PLANNING ALGORITHMS FOR CONTINUOUS MDPS WP2: PLANNING ALGORITHMS FOR STOCHASTIC GAMES WP3: FROM COMPLEX GAMES TO SIMPLER ONES WP4: MULTI-AGENT REINFORCEMENT LEARNING 1. On continuous-state MDPs w/ hierarchical information 2. Solving Dec-POMDPs as Sequential-Move Continuous-State Multi

Machine learning (ML) and artificial intelligence are rising themes of research for decades because they have been considered as one way to produce new algorithms for solving striking challenges such as language understanding, best advice finding, automatic signal processing, fraud detection. The ex

This project aims at building on recent progress in game playing and deep reinforcement learning to design a new generation of more general constraint satisfaction and optimization solvers. Upon completion, this project will contribute to make basic (or even naive) problem formulations far more e

Following the linked data principles, the Linked Open Data (LOD) promotes a vision of a global decentralized Knowledge Graph. However, the LOD KGs face serious technical and non-technical issues: The size of KGs is increased dramatically, raising issues on scalability, and the current metadata avail

The Internet of Things connects physical devices offering sensing or actuating with their vicinity. The ever-growing capabilities of devices allow to imagine new architectures including them as first class citizens. New added-value applications can then be envisioned in smart agriculture, smart buil

The SEMBA project researches the relations between the observations obtained with multimodal acquisition modalities (internal and external scanning), and provides innovative methods in order to infer the internal measurements from the dynamic external ones. Precisely, SEMBA has two main objectives.

Astronomical surveys planned for the coming years will produce data that present analysis challenges not only because of their scale (hundreds of petabytes), but also by the complexity of the measurement challenges on very deep images (for instance subpercent-level measurement of colors or shapes on

The growth and diversification in data collection techniques has led to tremendous changes in the optimization methods used in machine learning. Several research directions have recently proven to be able to scale up to current challenges. Among them, let us focus on two promising trends: i) Dimens

Business intelligence (BI) technologies such as data warehousing and On-Line Analysis Processing (OLAP) are major decision-support tools that used to necessitate heavy financial and human investment. Yet, there now exist numerous free BI suites including proprietary, open source and/or cloud solutio

ExTENSoR proposes fundamental research that aims to explore the potential of end-to-end and automatically learned / evolutive artificial neural networks for the automatic processing and classification of speech signals. ExTENSoR will investigate their use as an alternative to hand-crafted features a

There is a growing concern about misinformation or biased information in public communication, whether in traditional media or social forums. While automating fact-checking has received a lot of attention, the problem of fair information is actually much larger and includes more insidious forms lik

Neural networks are a fundamental tool for solving various artificial intelligence tasks, such as supervised and unsupervised classification. Recent progress is linked to deep neural networks with extremely large number of layers, which helped to achieve remarkable results in the context of many ap

It has recently become possible to capture time-varying 3D point clouds at high spatial and temporal resolution. This allows in particular for high-quality acquisitions of human bodies in motion. However, tools to process and analyze these data robustly and automatically are still missing. Such tool

Artificial Intelligence (AI) is of national strategic importance due to the impressive results of deep learning algorithms in different domains such as natural language processing (NLP), communications, medicine, law, political analytics, and military with a wide range of applications. France is bec

Recent technological advances rely on accurate decision support systems that have been constructed as black boxes. That is, the system's internal logic is not available to the user, e.g., due to the complexity of system. This lack of explanation can lead to technical, ethical, and legal issues. For

The main objective on this project is to provide an innovative framework for optimal control of stochastic distributed agents. Restless bandit allocation is one particular example where the control that can be sent to each arm is restricted to an on/off signal. The originality of this framework is t

Quantum Machine Learning is an emerging field of research, with fast growth. This research field is largely driven by the desire to develop artificial intelligence that uses quantum technologies to improve the speed and performance of learning algorithms. Strong interdisciplinary collaborations are

The proposed project lies in the field of computer vision and deep learning. We particularly study image classification and retrieval. As for machine learning, computer vision has witnessed a core change with the recent re-popularization of Deep Neural Networks (DNN). Even though the recent dee