Search

The demand for explainable and energy-efficient artificial intelligence (AI) systems for edge computing has led to significant interest in electronic systems dedicated to Bayesian inference. Traditional designs of such systems often rely on stochastic computing, which offers high energy efficiency but suffers from latency issues and struggles with low-probability values. In this paper, we introduce the logarithmic memristor-based Bayesian machine, an innovative design that leverages the unique properties of memristors and logarithmic computing as an alternative to stochastic computing. We present a prototype machine fabricated in a hybrid CMOS/hafnium-oxide memristor process. We validate the versatility and robustness of our system through experimental validation and extensive simulations in two distinct applications: gesture recognition and sleep stage classification. The logarithmic approach simplifies the computational model by converting multiplications into additions and enhances the handling of low-probability events, which are crucial in time-dependent tasks. Our results demonstrate that the logarithmic Bayesian machine achieves superior performance in terms of accuracy and energy efficiency compared to its stochastic counterpart, particularly in scenarios involving complex probabilistic models. This work paves the way for the deployment of advanced AI capabilities in edge devices, where power efficiency and reliability are paramount.

Serological tests are important for understanding the physiopathology and following the evolution of the Covid-19 pandemic. Assays based on flow cytometry (FACS) of tissue culture cells expressing the spike (S) protein of SARS-CoV-2 have repeatedly proven to perform slightly better than the plate-based assays ELISA and CLIA (chemiluminescent immuno-assay), and markedly better than lateral flow immuno-assays (LFIA). Here, we describe an optimized and very simple FACS assay based on staining a mix of two Jurkat cell lines, expressing either high levels of the S protein (Jurkat-S) or a fluorescent protein (Jurkat-R expressing m-Cherry, or Jurkat-G, expressing GFP, which serve as an internal negative control). We show that the Jurkat-S\&R-flow test has a much broader dynamic range than a commercial ELISA test and performs at least as well in terms of sensitivity and specificity. Also, it is more sensitive and quantitative than the hemagglutination-based test HAT, which we described recently. The Jurkat-flow test requires only a few microliters of blood; thus, it can be used to quantify various Ig isotypes in capillary blood collected from a finger prick. It can be used also to evaluate serological responses in mice, hamsters, cats and dogs. FACS tests offer a very attractive solution for laboratories with access to tissue culture and flow cytometry who want to monitor serological responses in humans or in animals, and how these relate to susceptibility to infection, or re-infection, by the virus, and to protection against Covid-19.

In recent years, a considerable research effort has shown the energy benefits of implementing neural networks with memristors or other emerging memory technologies. However, for extreme-edge applications with high uncertainty, access to reduced amounts of data, and where explainable decisions are required, neural networks may not provide an acceptable form of intelligence. Bayesian reasoning can solve these concerns, but it is computationally expensive and, unlike neural networks, does not translate naturally to memristor-based architectures. In this work, we introduce, demonstrate experimentally on a fully fabricated hybrid CMOS-memristor system, and analyze a Bayesian machine designed for highly-energy efficient Bayesian reasoning. The architecture of the machine is obtained by writing Bayes' law in a way making its implementation natural by the principles of distributed memory and stochastic computing, allowing the circuit to function using solely local memory and minimal data movement. Measurements on a fabricated small-scale Bayesian machine featuring 2,048 memristors and 30,080 transistors show the viability of this approach and the possibility of overcoming the challenges associated with its design: the inherent imperfections of memristors, as well as the need to distribute very locally higher-than-nominal supply voltages. The design of a scaled-up version of the machine shows its outstanding energy efficiency on a real-life gesture recognition task: a gesture can be recognized using 5,000 times less energy than using a microcontroller unit. The Bayesian machine also features several desirable features, e.g., instant on/off operation, compatibility with low supply voltages, and resilience to single-event upsets. These results open the road for Bayesian reasoning as an attractive way for energy-efficient, robust, and explainable intelligence at the edge.

Reconstruction of the tridimensional geometry of a visual scene using the binocular disparity information is an important issue in computer vision and mobile robotics, which can be formulated as a Bayesian inference problem. However, computation of the full disparity distribution with an advanced Bayesian model is usually an intractable problem, and proves computationally challenging even with a simple model. In this paper, we show how probabilistic hardware using distributed memory and alternate representation of data as stochastic bitstreams can solve that problem with high performance and energy efficiency. We put forward a way to express discrete probability distributions using stochastic data representations and perform Bayesian fusion using those representations, and show how that approach can be applied to diparity computation. We evaluate the system using a simulated stochastic implementation and discuss possible hardware implementations of such architectures and their potential for sensorimotor processing and robotics., Comment: Preprint of article submitted for publication in International Journal of Approximate Reasoning and accepted pending minor revisions

Living organisms survive and multiply even though they have uncertain and incomplete information about their environment and imperfect models to predict the consequences of their actions. Bayesian models have been proposed to face this challenge. Indeed, Bayesian inference is a way to do optimal reasoning when only uncertain and incomplete information is available. Various perceptive, sensory-motor, and cognitive functions have been successfully modeled this way. However, the biological mechanisms allowing animals and humans to represent and to compute probability distributions are not known. It has been proposed that neurons and assemblies of neurons could be the appropriate scale to search for clues to probabilistic reasoning. In contrast, in this paper, we propose that interacting populations of macromolecules and diffusible messengers can perform probabilistic computation. This suggests that probabilistic reasoning, based on cellular signaling pathways, is a fundamental skill of living organisms available to the simplest unicellular organisms as well as the most complex brains.

This paper advocates the exploration of the full state of recorded real-time strategy (RTS) games, by human or robotic players, to discover how to reason about tactics and strategy. We present a dataset of StarCraft games encompassing the most of the games' state (not only player's orders). We explain one of the possible usages of this dataset by clustering armies on their compositions. This reduction of armies compositions to mixtures of Gaussian allow for strategic reasoning at the level of the components. We evaluated this clustering method by predicting the outcomes of battles based on armies compositions' mixtures components, Comment: Artificial Intelligence in Adversarial Real-Time Games 2012, Palo Alto : United States (2012)

The task of keyhole (unobtrusive) plan recognition is central to adaptive game AI. "Tech trees" or "build trees" are the core of real-time strategy (RTS) game strategic (long term) planning. This paper presents a generic and simple Bayesian model for RTS build tree prediction from noisy observations, which parameters are learned from replays (game logs). This unsupervised machine learning approach involves minimal work for the game developers as it leverage players' data (com- mon in RTS). We applied it to StarCraft1 and showed that it yields high quality and robust predictions, that can feed an adaptive AI., Comment: 7 pages; Artificial Intelligence and Interactive Digital Entertainment Conference (AIIDE 2011), Palo Alto : \'Etats-Unis (2011)

This paper describes an application of Bayesian programming to the control of an autonomous avatar in a multiplayer role-playing game (the example is based on World of Warcraft). We model a particular task, which consists of choosing what to do and to select which target in a situation where allies and foes are present. We explain the model in Bayesian programming and show how we could learn the conditional probabilities from data gathered during human-played sessions., Comment: 30th international workshop on Bayesian Inference and Maximum Entropy, Chamonix : France (2010)

Computing and storing probabilities is a hard problem as soon as one has to deal with complex distributions over multiple 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 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.