Contexts in source publication

Context 1
... neurons, or nerve cells, are electrically excitable specialized cells that constitute the biggest part of the most advanced and complex organ in a mammalian body, the brain. Neurons mainly consist of three parts which are the cell body (or soma), the axon and dendrites (see Figure 2.1). Dendrites carry incoming electrical signals originating from other neurons while the axons deliver electrical signals to other individual target cells. A single neuron can be connected to as many as 10,000 other neurons. The junction that permits two neurons to exchange electrical signals is called synapse. One of the most remarkable characteristics of neurons is their ability to propagate electrical signals very fast and over very large distances. They achieve this by rapidly changing the difference in voltage between the interior and exterior parts, i.e. their membrane potential, which results the generation of electrical pulses that are called action potential or spikes. Artificial Neural Networks (ANN) are mathematical models inspired by the biological neural networks. Classical ANN models consist of mathematical substitutes of the processing elements of the real neural networks such as neurons, axons, dendrites and synapses (see Figure 2.2) as well as a specific topology i.e. the network architecture that connects the various types of neurons. The wide impact of ANN systems comes from their ability to learn. According to Mitchell [5], "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P , if its performance at tasks in T , as measured by P , improve by experience E." Likewise, ANNs can be configured in a way that the application of an input dataset results in the corresponding desired output. More specifically, the process of learning (or training) in classical ANNs can be described as the repeated adjustment of the synaptic weights between the neurons, according to a learning rule, while the system is being fed with input data, whose desired output is known. Then, the process terminates when the error between desired and the real output is ...
Context 2
... neurons, or nerve cells, are electrically excitable specialized cells that constitute the biggest part of the most advanced and complex organ in a mammalian body, the brain. Neurons mainly consist of three parts which are the cell body (or soma), the axon and dendrites (see Figure 2.1). Dendrites carry incoming electrical signals originating from other neurons while the axons deliver electrical signals to other individual target cells. A single neuron can be connected to as many as 10,000 other neurons. The junction that permits two neurons to exchange electrical signals is called synapse. One of the most remarkable characteristics of neurons is their ability to propagate electrical signals very fast and over very large distances. They achieve this by rapidly changing the difference in voltage between the interior and exterior parts, i.e. their membrane potential, which results the generation of electrical pulses that are called action potential or spikes. Artificial Neural Networks (ANN) are mathematical models inspired by the biological neural networks. Classical ANN models consist of mathematical substitutes of the processing elements of the real neural networks such as neurons, axons, dendrites and synapses (see Figure 2.2) as well as a specific topology i.e. the network architecture that connects the various types of neurons. The wide impact of ANN systems comes from their ability to learn. According to Mitchell [5], "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P , if its performance at tasks in T , as measured by P , improve by experience E." Likewise, ANNs can be configured in a way that the application of an input dataset results in the corresponding desired output. More specifically, the process of learning (or training) in classical ANNs can be described as the repeated adjustment of the synaptic weights between the neurons, according to a learning rule, while the system is being fed with input data, whose desired output is known. Then, the process terminates when the error between desired and the real output is ...
Context 3
... model is able mimic the dynamics of the different types of neocortical neurons almost as well as the Hodgkin-Huxley model and it supports the majority of the features shown in the table of Figure 2.3 which makes it one of the most biologically plausible choices. The only transparent weakness of Izhikevich model is the lack of biophysical meaning of the concepts used in the ...
Context 4
... above features were characterized as very important, in terms of AI research, mainly because there were not available with other proposed test-bets. The core of the GameBots 2004 system is a plug-in of Unreal Tournament 2004 that handles the communication between characters in the game and bot clients, via network sockets and TCP/IP protocol. (see Figure ...
Context 5
... this mode the GUI is arranged in a way that both information regarding the controlled avatar and the system that is used to control it, are visible. In particular, all the features of the visualization tool that have been described above are enabled, and the window that illustrates the environment looks like Figure 4.2. However, the user can disable any of the illustrated features in order to speed up the processing of demanding ...
Context 6
... workspace theory (GWT) is a simple and widely used cognitive architecture that consists of a model of consciousness and it was first suggested by Baars [8,9,7]. According to this theory, the most competitive part of cognitive content becomes conscious, which means that it becomes globally accessible to the rest parallel unconscious cognitive processes (see Figure 2.4). To support that view, GWT relies on the assumption that a large part of the primate brain consists of highly specialized regions. This can be considered consistent with the current human knowledge of the brain. ...
Context 7
... Hodgkin-Huxley is one of the oldest mathematical spiking neuron models, it is able to mimic the behaviour of different types of neurons very accurately. However, as shown in Figure 2.3, a significant drawback is its excessive computational cost. For the great importance of their work, Hodgkin and Huxley received the Nobel Prize in Physiology or Medicine in ...
Context 8
... the first step of the GameBots -client communication, the server sends sensory data for the visible world over the network connections. When this data is processed by the client and the next actions are decided, the latter sends back commands and GameBots undertakes to apply the desired actions to the character (e.g. move, shoot, talk, etc) To help agents to navigate to the environment and researchers to implement their algorithms more easily, GameBots provides to the connected clients some extra types of information under the corresponding requests, such as information related to path planning or adjustable ray casting. Figure 2.6: Gamebots architecture (human players connect directly to the UT server, and bots connect through the Gamebots Module) (Image adapted from ...
Context 9
... Tournament 2004 (UT2004) is a futuristic video game co-developed by Epic Games and Digital Extremes as a sequel of UT2003 and the original Unreal Tournament. It falls into a category of video games known as first-person shooters (FPS), where all real time players exist in a 3D virtual world (see Figure 2.5) with simulated physics and a variety of tools that give the players additional abilities. As implied by the term first person, the senses of every player are limited by their location, bearings, and occlusion within the virtual world. Like its predecessors, UT2004 was designed mostly as an head-to-head arena FPS, the so called 'Deathmatch' mode. However, it also includes two more core game types 'Capture the Flag' and 'Domination' as well as less emphasized modes and a significant number of maps. Also, active online Unreal Tournament's community has been consistently creating new maps and game types. Furthermore, UT2004 consist of fast, dynamic and complex 3D simulation engine, widely available at a very small cost and with a large user base. As a consequence of its popularity, UT2004 has been being tested by hundreds of thousands of ...

Similar publications

Article
Full-text available
Real-time accurate detection of three-dimensional (3D) objects is a fundamental necessity for self-driving vehicles. Most existing computer vision approaches are based on convolutional neural networks (CNNs). Although the CNN-based approaches can achieve high detection accuracy, their high energy consumption is a severe drawback. To resolve this pr...
Research
Full-text available
In this paper, precision agriculture system is introduced based on Wireless Sensor Network (WSN). Soil moisture considered one of environment factors that effect on crop. The period of irrigation must be monitored. Neural network capable of learning the behavior of the agricultural soil in absence of mathematical model. This paper introduced modifi...
Preprint
Full-text available
Spiking Neural Network (SNN), originating from the neural behavior in biology, has been recognized as one of the next-generation neural networks. Conventionally, SNNs can be obtained by converting from pre-trained Artificial Neural Networks (ANNs) by replacing the non-linear activation with spiking neurons without changing the parameters. In this w...
Conference Paper
Full-text available
Repetitive activities of daily living (ADL) and robotic active training are commonly practised in the rehabilitation of paralyzed patients, both of which have been proven rather effective to recover the locomotor function of impaired limbs. ADL classification based on electroencephalo-gram (EEG) is of great significance to perform active robotic re...
Preprint
Full-text available
Spiking neural networks (SNNs) are a promising candidate for biologically-inspired and energy efficient computation. However, their simulation is notoriously time consuming, and may be seen as a bottleneck in developing competitive training methods with potential deployment on neuromorphic hardware platforms. To address this issue, we provide an im...

Citations

... An An Artificial Neural Network Model[32] ...
Article
Full-text available
Modern technologies empower water distribution systems (WDS) for better services in the processes of water supply, storage, distribution, and recycling. They improve real-time monitoring, automating, and managing. However, the limitations of these technologies introduce cyber-physical attacks to the WDS. The main goals of cyber-physical attacks include disrupting normal operations and tampering the critical data, which have negative impacts on the WDS. Therefore, it is vital to develop and implement solutions to increase the security of the WDS by detecting and mitigating cyber-physical attacks. Since security for WDS is relatively new, there are no surveys on this topic in spite of its vital importance. Therefore, in this paper, we provide a comprehensive survey for the common cyber-physical attacks and common detection mechanisms for the WDS.We compare the attacks and detection methods with emphasis on ideas, methods, evaluation results, advantages, limitations, etc. We further provide a future research direction. We realize that there are still not much research attempts in this area and we hope that this work can trigger more research activities related to the WDS.
... Struktura jednog neurona je prilično jednostavna. Ona je prikazana na slici 1., a sastoji se od staničnog tijela (soma), aksona i niza dendrita[1]. Prednji kraj aksona je povezan s tijelom stanice dok se njegov zadnji kraj grana na završetke koji se nazivaju telodendroni i nastavljaju se na dendrite drugih neurona. ...
Conference Paper
Full-text available
In this paper a simplified model of an islanded type of microgrid using artificial neural networks is presented. The modelled microgrid consists of renewable energy source such as photovoltaic plant, battery storage system and a commercial building that represents an electric load. Each of the microgrid components is presented with a separate neural network in order to create a simplified model of a microgrid that approximates electrical behavior of a building.
... Details about the implementation of this architecture in a network of spiking neurons are given in Section IV. Because of space constraints, only a brief description is given of most of the specialized modules, with full details available in [43]. ...
Article
This paper describes the NeuroBot system, which uses a global workspace architecture, implemented in spiking neurons, to control an avatar within the Unreal Tournament 2004 (UT2004) computer game. This system is designed to display humanlike behavior within UT2004, which provides a good environment for comparing human and embodied AI behavior without the cost and difficulty of full humanoid robots. Using a biologically inspired approach, the architecture is loosely based on theories about the high-level control circuits in the brain, and it is the first neural implementation of a global workspace that has been embodied in a complex dynamic real-time environment. NeuroBot's humanlike behavior was tested by competing in the 2011 BotPrize competition, in which human judges play UT2004 and rate the humanness of other avatars that are controlled by a human or a bot. NeuroBot came a close second, achieving a humanness rating of 36%, while the most human human reached 67%. We also developed a humanness metric that combines a number of statistical measures of an avatar's behavior into a single number. In our experiments with this metric, NeuroBot was rated as 33% human, and the most human human achieved 73%.
... Details about the implementation of this architecture in a network of spiking neurons are given in Section IV. Because of space constraints, only a brief description is given of most of the specialized modules, with full details available in [43]. ...
Conference Paper
Full-text available
This paper describes a system that uses a global workspace architecture implemented in spiking neurons to control an avatar within the Unreal Tournament 2004 (UT2004) computer game. This system is designed to display human-like behaviour within UT2004, which provides a good environment for comparing human and embodied AI behaviour without the cost and difficulty of full humanoid robots. Using a biologically-inspired approach, the architecture is loosely based on theories about the high level control circuits in the brain, and it is the first neural implementation of a global workspace that is embodied in a dynamic real time environment. At its current stage of development the system can navigate through UT2004 and shoot opponents. We are currently completing the implementation and testing in preparation for the human-like bot competition at CIG 2011 in September.