• The Journal of the Acoustical Society of Korea
• /
• v.22 no.1E
• /
• pp.3-10
• /
• 2003

This paper describes two kinds of neural networks for stereoscopic vision, which have been applied to an identification of human speech. In speech recognition based on the stereoscopic vision neural networks (SVNN), the similarities are first obtained by comparing input vocal signals with standard models. They are then given to a dynamic process in which both competitive and cooperative processes are conducted among neighboring similarities. Through the dynamic processes, only one winner neuron is finally detected. In a comparative study, with, the average phoneme recognition accuracy on the two-layered SVNN was 7.7% higher than the Hidden Markov Model (HMM) recognizer with the structure of a single mixture and three states, and the three-layered was 6.6% higher. Therefore, it was noticed that SVNN outperformed the existing HMM recognizer in phoneme recognition.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.06a
• /
• pp.113-117
• /
• 1998

We propose a way to show the inherent learning complexity for the multilayer perceptron. We display the solution space and the error surfaces on the input space of a single neuron with two inputs. The evolution of its weights will follow one of the two error surfaces. We observe that when we use the back-propagation(BP) learning algorithm (1), the wight cam not jump to the lower error surface due to the implicit continuity constraint on the changes of weight. The self-relaxation approach is to explicity find out the best combination of all neurons' two error surfaces. The time complexity of training a multilayer perceptron by self-relaxationis exponential to the number of neurons.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.7
• /
• pp.1324-1330
• /
• 2007

In this paper, we propose pattern recognition algorithm for activities of daily living by adopting extreme learning machine based on single layer feedforward networks(SLFNs) to the signal from bidirectional accelerometer. For activity classification, 20 persons are participated and we acquire 6, types of signals at standing, walking, running, sitting, lying, and falling. Then, we design input vector using reduced model for ELM input. In ELM classification results, we can find accuracy change by increasing the number of hidden neurons. As a result, we find the accuracy is increased by increasing the number of hidden neuron. ELM is able to classify more than 80 % accuracy for experimental data set when the number of hidden is more than 20.

• International Journal of Control, Automation, and Systems
• /
• v.3 no.spc2
• /
• pp.302-307
• /
• 2005

A novel neuro controller based simple neuro-structure with modified error function is introduced in this paper. This controller consists of two independent controllers, known as the voltage regulator and the angular controller. The voltage regulator is used to modify terminal voltage for the purpose of tracking a reference voltage. The angular controller is utilized to guarantee the stability of the system. In this structure each neuron uses a linear hard limit activation function that depends on the controlled variable and its derivatives. There is no need for parameter identification or any off-line training data. Two proposed controllers are merged by a smooth switch to build a complete controller. The effectiveness of the proposed novel control action is demonstrated through some computer simulations on a Single-Machine Infinite-Bus (SMIB) power system.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.780-783
• /
• 2007

A Many studies have been made on the prediction of human voluntary movement intention in real-time based on invasive or non-invasive methods to help severely motor-disabled persons by offering some abilities of motor controls and communications. In the present study, we have developed an internet game driven by and/or linked to a brain-computer interface (BCI) system. Activities of two single neuronal units recorded from either hippocampus or prefrontal cortex of SD rats were used in real time to control two-dimensional movements of a robot, or a game object.

• BMB Reports
• /
• v.48 no.7
• /
• pp.388-394
• /
• 2015

The brain is an organ that consists of various cell types. As our knowledge of the structure and function of the brain progresses, cell type-specific research is gaining importance. Together with advances in sequencing technology and bioinformatics, cell type-specific transcriptome studies are providing important insights into brain cell function. In this review, we discuss 3 different cell type-specific transcriptome analyses i.e., Laser Capture Microdissection (LCM), Translating Ribosome Affinity Purification (TRAP)/RiboTag, and single cell RNA-Seq, that are widely used in the field of neuroscience. [BMB Reports 2015; 48(7): 388-394]

• Proceedings of the KOSOMBE Conference
• /
• v.1997 no.11
• /
• pp.31-34
• /
• 1997

A semiconductor microelectrode array has been successfully used in obtaining single unit recordings from medial giant nerve of clay fish, rat saphenous nerve and abdominal ganglia of aplysia. The recording device fabricated using silicon microfabrication techniques is a depth-probe type and, previously, has been mostly used to record from central nerve system of vertebrates. From invertebrates, and also from peripheral nerves of vertebrates, however, the quality of the recorded signal depends heavily on the recording conditions, such as the proximity of the electrode site to the nerve cells and the size of the neuron. We have modeled the signal to noise ratio as unctions of these parameters and compared the experimental data with the calculated values thus obtained.

• The Korean Journal of Physiology and Pharmacology
• /
• v.4 no.6
• /
• pp.427-437
• /
• 2000

The early studies of cardiac and smooth muscle cells provided evidence for two different calcium channels, the L-type (also called high-voltage activated [HVA]) and T-type (low-voltage activated [LVA]). These calcium channels provided calcium for muscle contractions and pace-making activities. As might be expected, the number of different calcium channels increased when researchers studied neurons and the identification of the neuronal calcium channels has proven to be much more difficult than with the muscle calcium channels. There are two reasons for this difficulty; (1) a larger number of different calcium channels in neurons and (2) many of the different calcium channels have similar kinetic properties. This review uses the N-type calcium channel to illustrate the difficulties in identifying and characterizing calcium channels in neurons. It shows that the discovery of toxins that can specifically block single calcium channel types has made it possible to easily and rapidly discern the physiological roles of the different calcium channels in the neuron, Without these toxins it is unlikely that progress would have been as rapid.

• The Transactions of the Korean Institute of Electrical Engineers B
• /
• v.51 no.12
• /
• pp.691-696
• /
• 2002

This paper presents a neural network based self-tuning fuzzy PID control scheme with variable learning rate for induction motor speed control. When induction motor is continuously used long time, its electrical and mechanical Parameters will change, which degrade the Performance of PID controller considerably. This Paper re-analyzes the fuzzy controller as conventional PID controller structure, introduces a single neuron with a back-propagation learning algorithm to tune the control parameters, and proposes a variable learning rate to improve the control performance. Proposed scheme is simple in structure and computational burden is small. The simulation using Matlab/Simulink and the experiment using dSPACE(DS1102) board are performed to verify the effectiveness of the proposed scheme.

• BMB Reports
• /
• v.57 no.4
• /
• pp.176-181
• /
• 2024

Memory allocation, which determines where memories are stored in specific neurons or synapses, has consistently been demonstrated to occur via specific mechanisms. Neuronal allocation studies have focused on the activated population of neurons and have shown that increased excitability via cAMP response element-binding protein (CREB) induces a bias toward memory-encoding neurons. Synaptic allocation suggests that synaptic tagging enables memory to be mediated through different synaptic strengthening mechanisms, even within a single neuron. In this review, we summarize the fundamental concepts of memory allocation at the neuronal and synaptic levels and discuss their potential interrelationships.