Bayesian reasoning is a machine learning approach that provides explainable outputs and excels in small-data situations with high uncertainty. However, it requires intensive memory access and computation and is, therefore, too energy-intensive for extreme edge contexts. Near-memory computation with memristors (or RRAM) can greatly improve the energy efficiency of its computations. Here, we report two fabricated integrated circuits in a hybrid CMOS-memristor process, featuring each sixteen tiny memristor arrays and the associated near-memory logic for Bayesian inference. One circuit performs Bayesian inference using stochastic computing, and the other uses logarithmic computation; these two paradigms fit the area constraints of near-memory computing well. On-chip measurements show the viability of both approaches with respect to memristor imperfections. The two Bayesian machines also operated well at low supply voltages. We also designed scaled-up versions of the machines. Both scaled-up designs can perform a gesture recognition task using orders of magnitude less energy than a microcontroller unit. We also see that if an accuracy lower than 86.9% is sufficient for this sample task, stochastic computing consumes less energy than logarithmic computing; for higher accuracies, logarithmic computation is more energy-efficient. These results highlight the potential of memristor-based near-memory Bayesian computing, providing both accuracy and energy efficiency.

International audience

International audience

This paper showcases the use of Bayesian models for real-time strategy (RTS) games AI in three distinct core components: micromanagement (units control), tactics (army moves and positions), and strategy (economy, technology, production, army types). The strength of having end-to-end probabilistic models is that distributions on specific variables can be used to interconnect different models at different levels of abstraction. We applied this modeling to StarCraft, and evaluated each model independently. Along the way, we produced and released a comprehensive data set for RTS machine learning.

This paper presents COSMO, a Bayesian computational model, which is expressive enough to carry out syllable production, perception and imitation tasks using motor, auditory or perceptuo-motor information.An imitation algorithm enables to learn the articulatory-to-acoustic mapping and the link between syllables and corresponding articulatory gestures, from acoustic inputs only: synthetic CV syllables generated with a human vocal tract model.We compare purely auditory, purely motor and perceptuo-motor syllable categorization under various noise levels.

In this paper, we present a Bayesian framework for the active multimodal perception of 3-D structure and motion. The design of this framework finds its inspiration in the role of the dorsal perceptual pathway of the human brain. Its composing models build upon a common egocentric spatial configuration that is naturally fitting for the integration of readings from multiple sensors using a Bayesian approach. In the process, we will contribute with efficient and robust probabilistic solutions for cyclopean geometry-based stereovision and auditory perception based only on binaural cues, modeled using a consistent formalization that allows their hierarchical use as building blocks for the multimodal sensor fusion framework. We will explicitly or implicitly address the most important challenges of sensor fusion using this framework, for vision, audition, and vestibular sensing. Moreover, interaction and navigation require maximal awareness of spatial surroundings, which, in turn, is obtained through active attentional and behavioral exploration of the environment. The computational models described in this paper will support the construction of a simultaneously flexible and powerful robotic implementation of multimodal active perception to be used in real-world applications, such as human-machine interaction or mobile robot navigation.

In this paper we consider the application of a naïve Bayes model for the evaluation of fraud risk connected with government agencies. This model applies probabilistic classifiers to support a generic risk assessment model, allowing for more efficient and effective use of resources for fraud detection in government transactions, and assisting audit agencies in transitioning from reactive to proactive fraud detection model.

This paper concerns the incremental learning of hierarchies of representations of space in artificial or natural cognitive systems. We propose a mathematical formalism for defining space representations (Bayesian Maps) and modelling their interaction in hierarchies of representations (sensorimotor interaction operator). We illustrate our formalism with a robotic experiment. Starting from a model based on the proximity to obstacles, we learn a new one related to the direction of the light source. It provides new behaviours, like phototaxis and photophobia. We then combine these two maps so as to identify parts of the environment where the way the two modalities interact is recognisable. This classification is a basis for learning a higher level of abstraction map that describes the large-scale structure of the environment. In the final model, the perception–action cycle is modelled by a hierarchy of sensorimotor models of increasing time and space scales, which provide navigation strategies of increasing complexities.

Domestic and real world robotics requires continuous learning of new skills and behaviors to interact with humans. Auto-supervised learning, a compromise between supervised and completely unsupervised learning, consist in relying on previous knowledge to acquire new skills. We propose here to realize auto-supervised learning by exploiting statistical regularities in the sensorimotor space of a robot. In our context, it corresponds to achieve feature selection in a Bayesian programming framework. We compare several feature selection algorithms and validate them on a real robotic experiment.

This paper deals with the probabilistic modeling of space, in the context of mobile robot navigation. We define a formalism called the Bayesian map, which allows incremental building of models, thanks to the superposition operator, which is a formally well-defined operator. Firstly, we present a syntactic version of this operator, and secondly, a version where the previously obtained model is enriched by experimental learning. In the resulting map, locations are the conjunction of underlying possible locations, which allows for more precise localization and more complex tasks. A theoretical example validates the concept, and hints at its usefulness for realistic robotic scenarios.

We are interested in probabilistic models of space and navigation. We describe an experiment where a Koala robot uses experimental data, gathered by randomly exploring the sensorimotor space, so as to learn a model of its interaction with the environment. This model is then used to generate a variety of new behaviors, from obstacle avoidance to wall following to ball pushing, which were previously unknown by the robot. The learned model can be seen as a building block for a hierarchical control architecture based on the Bayesian map formalism.

This work deals with the domain of space modeling for mobile robotics. It offers a comparison of probabilistic and biomimetic models of navigation. Both approaches are shown to be quite complementary: while the probabilistic methods exploit sound theoretical grounds, they lack the modularity and, as a consequence, flexibility, of their biomimetic counterparts. We propose a new formalism, called the Bayesian Map formalism, that attempts to bridge the gap between the two domains: it is based on Bayesian modeling and inference for defining the building blocks, and uses operators for building hierarchies of models.

This paper presents a new method for probabilistic modeling of space, called the Bayesian Map formalism. It offers a generalization of some common approaches found in the literature, as it does not constrain the dependency structure of the probabilistic model. The formalism allows incremental building of hierarchies of models, by the use of the Abstraction operator. In the resulting hierarchy, localization in the high level model is based on probabilistic competition of the lower level models. Experimental results validate the concept, and hint at its usefulness for large scale scenarios.

Autonomous navigation of a mobile robot along a predefined trajectory is a widely studied problem in the robotics community. We propose a Bayesian architecture that aims at being able to replay any sensori-motor trajectory trajectory defined as a sequence of perceptions and actions - as long as the robot starts in its neighbourhood. In order to increase robustness, we also use this Bayesian framework to estimate system self-confidence while the robot is moving. This work has been validated both on a simulated robot and on a real robot: the CyCab.

The recent development of a new kind of public transportation system relies on a particular double-steering kinematic structure enhancing maneuverability in cluttered environments such as downtown areas. We call bi-steerable car a vehicle showing this kind of kinematics. Endowed with autonomy capacities, the bi-steerable car ought to combine suitably and safely a set of abilities: simultaneous localisation and environment modelling, motion planning and motion execution amidst moderately dynamic obstacles. In this paper we address the integration of these four essential autonomy abilities into a single application. Specifically, we aim at reactive execution of planned motion. We address the fusion of controls issued from the control law and the obstacle avoidance module using probabilistic techniques.

A prerequisite to the design of future Advanced Driver Assistance Systems for cars is a sensing system providing all the information required for high-level driving assistance tasks. Carsense is a European project whose purpose is to develop such a new sensing system. It will combine different sensors (laser, radar and video) and will rely on the fusion of the information coming from these sensors in order to achieve better accuracy, robustness and an increase of the information content. This paper demonstrates the interest of using probabilistic reasoning techniques to address this challenging multi-sensor data fusion problem. The approach used is called Bayesian Programming. It is a general approach based on an implementation of the Bayesian theory. It was introduced first to design robot control programs but its scope of application is much broader and it can be used whenever one has to deal with problems involving uncertain or incomplete knowledge.

Computing and storing probabilities is a hard problem as soon as one has to deal with complex distributions over multiples random variables. The problem of efficient representation of probability distributions is central in term of computational efficiency in the field of probabilistic reasoning. The main problem arises when dealing with joint probability distributions over a set of random variables: they are always represented using huge probability arrays. In this paper, a new method based on a binary-tree representation is introduced in order to store efficiently very large joint distributions. Our approach approximates any multidimensional joint distributions using an adaptive discretization of the space. We make the assumption that the lower is the probability mass of a particular region of feature space, the larger is the discretization step. This assumption leads to a very optimized representation in term of time and memory. The other advantages of our approach are the ability to refine dynamically the distribution every time it is needed leading to a more accurate representation of the probability distribution and to an anytime representation of the distribution.

Evolve safely in an unchanged environment and possibly following an optimal trajectory is one big challenge presented by situated robotics research field. Collision avoidance is a basic security requirement and this paper proposes a solution based on a probabilistic approach called Bayesian Programming. This approach aims to deal with the uncertainty, imprecision and incompleteness of the information handled. Some examples illustrate the process of embodying the programmer preliminary knowledge into a Bayesian program and experimental results of these examples implementation in an electrical vehicle are described and commented. Some videos illustrating these experiments can be found at http://www-laplace.imag.fr.

More and more fields of applied computer science involve fusion of multiple data sources, such as sensor readings or model decision. However, incompleteness of the model prevents the programmer from having an absolute precision over their variables. Therefore Bayesian framework can be adequate fro such a process as it allows handling of uncertainty. We will be interested in the ability to express any fusion process as a product, for it can lead to reduction of complexity in time and space. We study in this paper various fusion schemes and propose to add consistency variable to justify the use of a product to compute distribution over the fused variable. We will then show application of this new fusion process to localization of a mobile robot and obstacle avoidance.

A prerequisite to the design of future advanced driver assistance systems for cars is a sensing system that provides all the information required for high-level driving assistance tasks. Carsense is a European project whose purpose is to develop such a new sensing system. It combines different sensors (laser, radar and video) and relies on the fusion of the information coming from these sensors in order to achieve better accuracy, robustness and an increase of the information content. This paper demonstrates the interest of using probabilistic reasoning techniques to address this challenging multi-sensor data fusion problem. The approach used is called Bayesian programming. It is a general approach based on an implementation of the Bayesian theory. It was introduced initially to design robot control programs but its scope of application including uncertain or incomplete knowledge handling problems.

This paper describes how a mobile robot (a six-wheeled Koala equipped with a PAL pan-tilt camera) can chase an elusive target (a remote controlled toy car) in a unknown and unconstrained environment. First, the paper demonstrates the efficiency, simplicity, and adequacy of Bayesian robot programming to quickly develop such applications. Next, it illustrates that a high information compression ratio may be obtained by some pertinent sensory-motor decoupling.

We present a Bayesian CAD system for robotic applications. We address the problem of the propagation of geometric uncertainties, and how to take this propagation into account when solving inverse problems. We describe the methodology we use to represent and handle uncertainties using probability distributions of the system's parameters and sensor measurements. It may be seen as a generalization of constraint-based approaches where we express a constraint as a probability distribution instead of a simple equality or inequality. Appropriate numerical algorithms used to apply this methodology are also described. Using an example, we show how to apply our approach by providing simulation results using our CAD system.

We present a new approach to path planning, called the ``Ariadne's clew algorithm''. It is designed to find paths in high-dimensional continuous spaces and applies to robots with many degrees of freedom in static, as well as dynamic environments --- ones where obstacles may move. The Ariadne's clew algorithm comprises two sub-algorithms, called SEARCH and EXPLORE, applied in an interleaved manner. EXPLORE builds a representation of the accessible space while SEARCH looks for the target. Both are posed as optimization problems. We describe a real implementation of the algorithm to plan paths for a six degrees of freedom arm in a dynamic environment where another six degrees of freedom arm is used as a moving obstacle. Experimental results show that a path is found in about one second without any pre-processing.

A computational approach to direct and generalized inverse model acquisition is presented. The approach is based on a proposed method to direct model acquisition from partial information. The method decomposes a hyper-space function in one variable functions, simplifying the learning problem. The acquired direct model is then implemented in a tree-like structure that can be used in the inverse sense without additional learning effort. The authors' approach is able to acquire complete models in hyper-spaces requiring only selected data focused in one-dimension sub-spaces, strongly reducing the data acquisition effort. The authors' approach is particularly interesting for applications in robotics. The acquisition of direct models in robotics frequently takes place in high dimension phase spaces. When traditional approximation methods are used, enormous data bases, containing the examples to be interpolated, are required.

The goal of the work described is to build a path planner able to drive a robot in a dynamic environment where the obstacles are moving. In order to do so, the authors propose a method, called Ariadne's clew algorithm, to build a global path planner based on the combination of two local planning algorithms: an explore algorithm and a search algorithm. The purpose of the explore algorithm is to collect information about the environment with an increasingly fine resolution by placing landmarks in the searched space. The goal of the search algorithm is to opportunistically check if the target can be easily reached from any given placed landmark. The Ariadne's clew algorithm is shown to be very fast is most cases, allowing planning in dynamic environment. It is shown to be complete, which means that it is sure to find a path when one exists. A massively parallel implementation of this algorithm is described.

Article A parallel genetic algorithm for the graph partitioning problem Share on Authors: E.-G. Talbi Laboratoire de Génie Informatique / Institut IMAG, University of Grenoble Laboratoire de Génie Informatique / Institut IMAG, University of GrenobleView Profile , P. Bessière BP53X, F-38041 Grenoble, France and Laboratoire de Génie Informatique / Institut IMAG, University of Grenoble BP53X, F-38041 Grenoble, France and Laboratoire de Génie Informatique / Institut IMAG, University of GrenobleView Profile Authors Info & Claims ICS '91: Proceedings of the 5th international conference on SupercomputingJune 1991 Pages 312–320https://doi.org/10.1145/109025.109102Online:01 June 1991Publication History 39citation1,240DownloadsMetricsTotal Citations39Total Downloads1,240Last 12 Months29Last 6 weeks2 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access