Inhaltsverzeichnis
Auswahl der wissenschaftlichen Literatur zum Thema „Neurones à impulsions“
Geben Sie eine Quelle nach APA, MLA, Chicago, Harvard und anderen Zitierweisen an
Machen Sie sich mit den Listen der aktuellen Artikel, Bücher, Dissertationen, Berichten und anderer wissenschaftlichen Quellen zum Thema "Neurones à impulsions" bekannt.
Neben jedem Werk im Literaturverzeichnis ist die Option "Zur Bibliographie hinzufügen" verfügbar. Nutzen Sie sie, wird Ihre bibliographische Angabe des gewählten Werkes nach der nötigen Zitierweise (APA, MLA, Harvard, Chicago, Vancouver usw.) automatisch gestaltet.
Sie können auch den vollen Text der wissenschaftlichen Publikation im PDF-Format herunterladen und eine Online-Annotation der Arbeit lesen, wenn die relevanten Parameter in den Metadaten verfügbar sind.
Zeitschriftenartikel zum Thema "Neurones à impulsions"
Kravchenko, S. V. „DEVELOPING A SYSTEM FOR NEURAL PROTOTYPING OF NEURAL PROSTHESES BASED ON THE HYBRID SOFTWARE AND HARDWARE IMPLEMENTATION OF SPIKING NEURAL NETWORKS“. PROCEEDINGS IN CYBERNETICS 22, Nr. 4 (2023): 26–32. http://dx.doi.org/10.35266/1999-7604-2023-4-4.
Der volle Inhalt der QuelleDissertationen zum Thema "Neurones à impulsions"
Godin, Christelle. „Contributions à l'embarquabilité et à la robustesse des réseaux de neurones en environnement radiatif : apprentissage constructif : neurones à impulsions“. École nationale supérieure de l'aéronautique et de l'espace (Toulouse ; 1972-2007), 2000. http://www.theses.fr/2000ESAE0013.
Der volle Inhalt der QuelleSoula, Hédi. „Dynamique et plasticité dans les réseaux de neurones à impulsions : étude du couplage temporel réseau / agent / environnement“. Lyon, INSA, 2005. http://theses.insa-lyon.fr/publication/2005ISAL0056/these.pdf.
Der volle Inhalt der QuelleAn «artificial life » approach is conducted in order to assess the neural basis of behaviours. Behaviour is the consequence of a good concordance between the controller, the agent’s sensori-motors capabilities and the environment. Within a dynamical system paradigm, behaviours are viewed as attractors in the perception/action space – derived from the composition of the internal and external dynamics. Since internal dynamics is originated by the neural dynamics, learning behaviours therefore consists on coupling external and internal dynamics by modifying network’s free parameters. We begin by introducing a detailed study of the dynamics of large networks of spiking neurons. In spontaneous mode (i. E. Without any input), these networks have a non trivial functioning. According to the parameters of the weight distribution and provided independence hypotheses, we are able to describe completely the spiking activity. Among other results, a bifurcation is predicted according to a coupling factor (the variance of the distribution). We also show the influence of this parameter on the chaotic dynamics of the network. To learn behaviours, we use a biologically plausible learning paradigm – the Spike-Timing Dependent Plasticity (STDP) that allows us to couple neural and external dynamics. Applying shrewdly this learning law enables the network to remain “at the edge of chaos” which corresponds to an interesting state of activity for learning. In order to validate our approach, we use these networks to control an agent whose task is to avoid obstacles using only the visual flow coming from its linear camera. We detail the results of the learning process for both simulated and real robotics platform
Soula, Hédi Favrel Joel Beslon Guillaume. „Dynamique et plasticité dans les réseaux de neurones à impulsions étude du couplage temporel réseau / agent / environnement /“. Villeurbanne : Doc'INSA, 2005. http://docinsa.insa-lyon.fr/these/pont.php?id=soula.
Der volle Inhalt der QuelleWu, Jiaming. „A modular dynamic Neuro-Synaptic platform for Spiking Neural Networks“. Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPASP145.
Der volle Inhalt der QuelleBiological and artificial neural networks share a fundamental computational unit: the neuron. These neurons are coupled by synapses, forming complex networks that enable various functions. Similarly, neuromorphic hardware, or more generally neuro-computers, also require two hardware elements: neurons and synapses. In this work, we introduce a bio-inspired spiking Neuro-Synaptic hardware unit, fully implemented with conventional electronic components. Our hardware is based on a textbook theoretical model of the spiking neuron, and its synaptic and membrane currents. The spiking neuron is fully analog and the various models that we introduced are defined by their hardware implementation. The neuron excitability is achieved through a memristive device made from off-the-shelf electronic components. Both synaptic and membrane currents feature tunable intensities and bio-mimetic dynamics, including excitatory and inhibitory currents. All model parameters are adjustable, allowing the system to be tuned to bio-compatible timescales, which is crucial in applications such as brain-machine interfaces. Building on these two modular units, we demonstrate various basic neural network motifs (or neuro-computing primitives) and show how to combine these fundamental motifs to implement more complex network functionalities, such as dynamical memories and central pattern generators. Our hardware design also carries potential extensions for integrating oxide-based memristors (which are widely studied in material science),or porting the design to very large-scale integration (VLSI) to implement large-scale networks. The Neuro-Synaptic unit can be considered as a building block for implementing spiking neural networks of arbitrary geometry. Its compact and modular design, as well as the wide availability of ordinary electronic components, makes our approach an attractive platform for building neural interfaces in medical devices, robotics, and artificial intelligence systems such as reservoir computing
Lorrain, Vincent. „Etude et conception de circuits innovants exploitant les caractéristiques des nouvelles technologies mémoires résistives“. Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS182/document.
Der volle Inhalt der QuelleIn this thesis, we study the dedicated computational approaches of deep neural networks and more particularly the convolutional neural networks (CNN).We highlight the convolutional neural networks efficiency make them interesting choice for many applications. We study the different implementation possibilities of this type of networks in order to deduce their computational complexity. We show that the computational complexity of this type of structure can quickly become incompatible with embedded resources. To address this issue, we explored differents models of neurons and architectures that could minimize the resources required for the application. In a first step, our approach consisted in exploring the possible gains by changing the model of neurons. We show that the so-called spiking models theoretically reduce the computational complexity while offering interesting dynamic properties but require a complete rethinking of the hardware architecture. We then proposed our spiking approach to the computation of convolutional neural networks with an associated architecture. We have set up a software and hardware simulation chain in order to explore the different paradigms of computation and hardware implementation and evaluate their suitability with embedded environments. This chain allows us to validate the computational aspects but also to evaluate the relevance of our architectural choices. Our theoretical approach has been validated by our chain and our architecture has been simulated in 28 nm FDSOI. Thus we have shown that this approach is relatively efficient with interesting properties of scaling, dynamic precision and computational performance. In the end, the implementation of convolutional neural networks using spiking models seems to be promising for improving the networks efficiency. Moreover, it allows improvements by the addition of a non-supervised learning type STDP, the improvement of the spike coding or the efficient integration of RRAM memory
Faouzi, Johann. „Machine learning to predict impulse control disorders in Parkinson's disease“. Electronic Thesis or Diss., Sorbonne université, 2020. http://www.theses.fr/2020SORUS048.
Der volle Inhalt der QuelleImpulse control disorders are a class of psychiatric disorders characterized by impulsivity. These disorders are common during the course of Parkinson's disease, decrease the quality of life of subjects, and increase caregiver burden. Being able to predict which individuals are at higher risk of developing these disorders and when is of high importance. The objective of this thesis is to study impulse control disorders in Parkinson's disease from the statistical and machine learning points of view, and can be divided into two parts. The first part consists in investigating the predictive performance of the altogether factors associated with these disorders in the literature. The second part consists in studying the association and the usefulness of other factors, in particular genetic data, to improve the predictive performance
Jouni, Zalfa. „Analog spike-based neuromorphic computing for low-power smart IoT applications“. Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPAST114.
Der volle Inhalt der QuelleAs the Internet of Things (IoT) expands with more connected devices and complex communications, the demand for precise, energy-efficient localization technologies has intensified. Traditional machine learning and artificial intelligence (AI) techniques provide high accuracy in radio-frequency (RF) localization but often at the cost of greater complexity and power usage. To address these challenges, this thesis explores the potential of neuromorphic computing, inspired by brain functionality, to enable energy-efficient AI-based RF localization. It introduces an end-to-end analog spike-based neuromorphic system (RF NeuroAS), with a simplified version fully implemented in BiCMOS 55 nm technology. RF NeuroAS is designed to identify source positions within a 360-degree range on a two-dimensional plane, maintaining high resolution (10 or 1 degree) even in noisy conditions. The core of this system, an analog-based spiking neural network (A-SNN), was trained and tested on a simulated dataset (SimLocRF) from MATLAB and an experimental dataset (MeasLocRF) from anechoic chamber measurements, both developed in this thesis.The learning algorithms for A-SNN were developed through two approaches: software-based deep learning (DL) and bio-plausible spike-timing-dependent plasticity (STDP). RF NeuroAS achieves a localization accuracy of 97.1% with SimLocRF and 90.7% with MeasLoc at a 10-degree resolution, maintaining high performance with low power consumption in the nanowatt range. The simplified RF NeuroAS consumes just over 1.1 nW and operates within a 30 dB dynamic range. A-SNN learning, via DL and STDP, demonstrated performance on XOR and MNIST problems. DL depends on the non-linearity of post-layout transfer functions of A-SNN's neurons and synapses, while STDP depends on the random noise in analog neuron circuits. These findings highlight advancements in energy-efficient IoT through neuromorphic computing, promising low-power smart edge IoT breakthroughs inspired by brain mechanisms
Spyrou, Theofilos. „Functional safety and reliability of neuromorphic computing systems“. Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS118.
Der volle Inhalt der QuelleThe recent rise of Artificial Intelligence (AI) has found a wide range of applications essentially integrating it gaining more and more ground in almost every field of our lives. With this steep integration of AI, it is reasonable for concerns to arise, which need to be eliminated before the employment of AI in the field, especially in mission- and safety-critical applications like autonomous vehicles. Spiking Neural Networks (SNNs), although biologically inspired, inherit only partially the remarkable fault resilience capabilities of their biological counterparts, being vulnerable to electronic defects and faults occurring at hardware level. Hence, a methodological exploration of the dependability characteristics of AI hardware accelerators and neuromorphic platforms is of utmost importance. This thesis tackles the subjects of testing and fault tolerance in SNNs and their neuromorphic implementations on hardware
Molino, Minero Erik. „Aportaciones a la identificación de señales impulsivas generadas por impactos“. Doctoral thesis, Universitat Politècnica de Catalunya, 2010. http://hdl.handle.net/10803/6374.
Der volle Inhalt der QuelleEl desarrollo principal de esta tesis se enfoca al problema de cómo compensar o reducir los efectos de dicha distorsión. Para ello, se han investigado y desarrollado los siguientes puntos:
1) Estudio de la teoría mecánica del impacto y desarrollo de un modelo matemático del proceso de impacto entre dos cuerpos rígidos. A través de este estudio se investigan las características de las señales impulsivas generadas por colisiones.
2) Definición de una metodología experimental para generar impactos repetibles y determinar los parámetros del modelo matemático. La metodología se sustenta en el diseño e implementación de un prototipo experimental para generar impactos controlados, entre un objeto de prueba y un impactor sensorizado. Para realizar los experimentos se han seleccionado como objetos de prueba un conjunto de cilindros, de aluminio, acero, bronce y latón, en distintos tamaños. Mediante un minucioso estudio y cálculo de los parámetros experimentales, se ha comprobado la validez del modelo matemático.
3) Estudio del problema de la medición indirecta y propuesta de un método de procesado de señales, basado en redes neuronales artificiales, para determinar un filtro inverso que permite estimar la señal del impacto (la fuerza del impacto en función del tiempo). Esta metodología adapta el proceso de entrenamiento a las características de las señales impulsivas que se generan durante una colisión, y que se han identificado a través del estudio y modelado del proceso de impacto. El entrenamiento hace uso de señales reales, que provienen de impactos experimentales generados a distintas velocidades de impacto, y de señales generadas por el modelo matemático.
4) Propuesta de una metodología para estimar el tipo de material y la masa de los objetos de prueba que colisionan. La problemática que se encuentra en este análisis radica en que tanto los objetos como sus respuestas, tienen características similares. Con el método que se propone en este trabajo de tesis, se busca identificar de forma correcta las características de los objetos. El procedimiento considera la extracción de parámetros de las señales vibratorias de los objetos y del uso de redes neuronales para identificar las respuestas.
5) Proceso de evaluación experimental de los métodos propuestos. Para determinar la validez de los métodos de procesado antes descrito, primero se analizan los sensores más adecuados para este tipo de señales, que al ser de muy corta duración tienen un ancho de banda muy grande. En segundo lugar, se ha implementado un sistema medición y adquisición para señales impulsivas.
Los resultados obtenidos muestran la validez de los métodos propuestos. Con respecto al modelo, se ha verificado su validez con los datos de los distintos objetos de prueba. Asimismo, se ha comprobado que con las señales experimentales, también de los distintos objetos de prueba, el método propuesto para mitigar la distorsión debida a la medición indirecta opera de forma correcta. De la misma forma, el método propuesto para identificar el tipo de material y la masa de los objetos, ha generado resultados satisfactorios.
In this thesis, the processing of impulsive signals generated by impacts between rigid bodies is investigated. One of the problems found when working with impacts is that their analysis is generally limited to indirect measurements: because collisions do not develop directly on the sensor, or it is not possible to install the sensor on the colliding bodies. This means that between the sensor and the point of impact there is a propagation medium that distorts the measured signal.
The main effort of this thesis focuses on the problem of how to compensate or to reduce the effects of such distortion. To do this, the following points have been investigated and developed:
1) The study of the mechanical impact theory and the development of a mathematical model of the impact process between two rigid bodies. Through this study, the characteristics of the impulsive signals generated by collisions are investigated.
2) Definition of an experimental methodology for generating repeatable impacts and for determining the parameters of the mathematical model. The methodology is based on the design and implementation of an experimental prototype for generating controlled impacts between a test object and a sensorized impactor. To perform the experiments, a set of different test objects have been selected, cylinders made form aluminum, steel, bronze and brass in different sizes. Through a careful study and calculation of the experimental parameters, the validity of the mathematical model has been verified.
3) Study of the indirect measurement problem, and proposal of a signal processing method, based on artificial neural networks, to determine an inverse filter in order to estimate the impacting signal (the impact force as a function of the time). This methodology adapts the training process to the characteristics of the impulsive signals that are generated during a collision, and that have been identified through the study and modeling of the impact process. The training uses real signals, which come from experimental impacts generated at different impacting velocities, and signals generated by a mathematical model of the impacting force.
4) Proposal for a methodology to estimate the type of material and mass of test objects that collide. The problem found in this analysis is that both, the objects and their responses, have similar characteristics. With the method proposed in this thesis, it is possible to identify correctly the characteristics of one of the objects. The procedure considers the extraction of parameters from the vibrating signals of the objects, and then uses a neural network to classify those parameters.
5) Evaluation process of the proposed methods. To determine the validity of the processing methods described above, first, the selection of the most appropriate sensors to acquire these signals has been analyzed (this signals have a very short duration and very large bandwidth). Secondly, a measurement and acquisition system for impulsive signals has been implemented.
The experimental results show the validity of the proposed methods. In the case of the model, its validity has been verified with data from different test objects, made from different materials. Also, the proposed method used to deal with the distortion due to the indirect measurement has been tested with experimental data, from impacts with different test objects, and the results show that it operates properly. Likewise, the proposed method to identify the type of material and mass of the test objects has generated satisfactory results.
Thiele, Johannes C. „Deep learning in event-based neuromorphic systems“. Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS403/document.
Der volle Inhalt der QuelleInference and training in deep neural networks require large amounts of computation, which in many cases prevents the integration of deep networks in resource constrained environments. Event-based spiking neural networks represent an alternative to standard artificial neural networks that holds the promise of being capable of more energy efficient processing. However, training spiking neural networks to achieve high inference performance is still challenging, in particular when learning is also required to be compatible with neuromorphic constraints. This thesis studies training algorithms and information encoding in such deep networks of spiking neurons. Starting from a biologically inspired learning rule, we analyze which properties of learning rules are necessary in deep spiking neural networks to enable embedded learning in a continuous learning scenario. We show that a time scale invariant learning rule based on spike-timing dependent plasticity is able to perform hierarchical feature extraction and classification of simple objects of the MNIST and N-MNIST dataset. To overcome certain limitations of this approach we design a novel framework for spike-based learning, SpikeGrad, which represents a fully event-based implementation of the gradient backpropagation algorithm. We show how this algorithm can be used to train a spiking network that performs inference of relations between numbers and MNIST images. Additionally, we demonstrate that the framework is able to train large-scale convolutional spiking networks to competitive recognition rates on the MNIST and CIFAR10 datasets. In addition to being an effective and precise learning mechanism, SpikeGrad allows the description of the response of the spiking neural network in terms of a standard artificial neural network, which allows a faster simulation of spiking neural network training. Our work therefore introduces several powerful training concepts for on-chip learning in neuromorphic devices, that could help to scale spiking neural networks to real-world problems