• IEIE Transactions on Smart Processing and Computing
• /
• v.4 no.2
• /
• pp.110-114
• /
• 2015

This paper reports a novel object-detection technique in video sequences. The proposed algorithm consists of detection of objects in illuminated and occluded videos by using object features and a neural network technique. It consists of two functional modules: region-based object feature extraction and continuous detection of objects in video sequences with region features. This scheme is proposed as an enhancement of the Lowe's scale-invariant feature transform (SIFT) object detection method. This technique solved the high computation time problem of feature generation in the SIFT method. The improvement is achieved by region-based feature classification in the objects to be detected; optimal neural network-based feature reduction is presented in order to reduce the object region feature dataset with winner pixel estimation between the video frames of the video sequence. Simulation results show that the proposed scheme achieves better overall performance than other object detection techniques, and region-based feature detection is faster in comparison to other recent techniques.

• Electronics and Telecommunications Trends
• /
• v.34 no.6
• /
• pp.87-99
• /
• 2019

Reinforcement learning is a type of machine learning paradigm that forces agents to repeat the observation-action-reward process to assess and predict the values of possible future action sequences. This allows the agents to incrementally reinforce the desired behavior for a given observation. Thanks to the recent advancements of deep learning, reinforcement learning has evolved into deep reinforcement learning that introduces promising results in various control and optimization domains, such as games, robotics, autonomous vehicles, computing, industrial control, and so on. In addition to this trend, a number of programming libraries have been developed for importing deep reinforcement learning into a variety of applications. In this article, we briefly review and summarize 10 representative deep reinforcement learning libraries and compare them from a development project perspective.

• International Journal of Control, Automation, and Systems
• /
• v.1 no.1
• /
• pp.76-82
• /
• 2003

This paper establishes some integral equalities formulated by zeros located in the convergence region of a Laplace transformable function. Using the definition of the Laplace transform, it shows that Laplace transformable functions have to satisfy the integral equalities in the time-domain, which can be applied to the understanding of the fundamental limitations on the control system represented by the transfer function. In the unity-feedback control scheme, another integral equality is derived on the output response of the system with open-loop poles located in the convergence region of the output function. From these integral equalities, two sufficient conditions related to undershoot and overshoot phenomena in the step response, respectively, are investigated.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.794-799
• /
• 2004

Recent paradigm in technology shifts from object-based technology to environment-based technology. Issue here is interaction among humans and the world around humans, which is natural and artificial "space." Holistic interactions based on "Mom (embodiment)" suggest a good starting point for exploring this issue. Soft engineering, "Mom," holistic interactions, tangible space, ubiquitous computing, science of emotion, and interactive media are key concepts in interactive technology. Interactive tangible media "iT_Media" is proposed to explore and synthesize these ideas. Interactive technology initiative (ITI) is an interdisciplinary research group to search for the proper technology and the proper way of implementing technology: "interactive technology" or "soft engineering." Some experimental activities conducted by ITI are presented in this session, "Interactive Technology."

• Transactions on Control, Automation and Systems Engineering
• /
• v.3 no.4
• /
• pp.242-249
• /
• 2001

In this paper, the open-loop state response of the two-time-scale systems by unified approach using the $\delta$-operator is presented with an example of the aircraft longitudinal dynamics. First, the $\delta$-operator system unifies both the continuous system and the discrete system simultaneously, and the $\delta$-operator approach improves the finite word-length characteristics. This saves more computing time than that of the discrete system. Second, the singular perturbation method by block diagonalization reduces the sizes and orders of the systems. This also reduces the floating-point operations (flops). The advantage of those two approaches is shown by comparing our results with the earlier ones in the illustrative example of the longitudinal motion of F-8 aircraft.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.5
• /
• pp.559-566
• /
• 2006

In the real system application, the interacting multiple model (IMM) based algorithm requires less computing resources as well as a good performance with respect to the various target maneuverings. And it further requires an easy design procedure in terms of its structures and parameters. To solve these problems, a fuzzy interacting multiple model (FIMM) algorithm, which is based on the basis sub-models defined by considering the maneuvering property and the time-varying mode transition probabilities designed by using the mode probabilities as inputs of a fuzzy decision maker, is proposed. To verify the performance of the proposed algorithm, airborne target tracking is performed. Simulation results show that the FIMM algorithm solves all problems in the real system application of the IMM based algorithm.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1263-1266
• /
• 1997

In a fault-tolerant modern manufacturing systms characterized by the configuration, in which automated redundant machines prone to unexpected failures are interconnected with other complex subsystems such as AGV's, robots, computer control systems to produce complete parts, faulures together with repairs and reconfigurations should be considered as the three basic events to be modeled for computing the performance of manufacturing systems. In this papre, transient analysis is applied to modular cell manufacturing systems form a performability viewpoint whose modeling adantage is that various performanc e measures can be evaluated compositely in the context of application. The hypothertical modular cells are modeled firstly with hybrid decomposition method and availability measures as special cases of performability are computed and comments on performabililty modeling analysis are mentioned.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.579-581
• /
• 2004

We wish to develop micro-total analysis system (TAS) on a chip, and to make a trial approach to solve the important problem that is to detect ions separated by the electric field. We propose an idea, which is as for rotational motions of dipolar ions, which are affected by the ion atmosphere in outer regions. This is a new kind of the ion-sensitive field effect transistor (ISFET). We wish to develop the ISFET chips, and give more effective, fast and sensitive, capillary electrophoresis is designed in near future.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.632-637
• /
• 1992

Recently many dynamics control algorithms using robot dynamic equation have been proposed. One of them, Kawato's feedback error learning scheme requires neither an accurate model nor parameter estimation and makes the robot motion closer to the desired trajectory by repeating operation. In this paper, the feedback error learning algorithm is implemented to control a robot system, 5 DOF revolute type movemaster. For this purpose, an actuator dynamic model is constructed considering equivalent robot dynamics model with respect to actuator as well as friction model. The command input acquired from the actuator dynamic model is the sum of products of unknown parameters and known functions. To compute the control algorithm, a parallel processing computer, transputer, is used and real-time computing is achieved. The experiment is done for the three major link of movemaster and its result is presented.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10b
• /
• pp.1440-1445
• /
• 1990

The system with softfware packages for control system design unifying and encompassing rule based control and conventional control based on numerical models were developed. Users who are not familiar with control theory, numerical computing, and artificial intelligence (AI) can perform system analysis, control design and development of AI control system without difficulty.