• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.2C
• /
• pp.181-185
• /
• 2009

When we classify underwater transient signals with frame-by-frame decision, a database design method for reference feature vectors influences on the system performance such as size of database, computational burden and recognition rate. In this paper the LBG vector quantization algorithm is applied to reduction of the number of feature vectors for each reference signal for efficient classification of underwater transient signals. Experimental results have shown that drastic reduction of the database size can be achieved while maintaining the classification performance by using the LBG vector quantization.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.6
• /
• pp.612-618
• /
• 2014

A pattern classifier-based approach for recognizing internal states of human participants in interactions is presented along with its experimental results. The approach includes a step for collecting video recordings of human-human interactions or humanrobot interactions and subsequently analyzing the videos based on human coded annotations. The annotation includes social signals directly observed in the video recordings and the internal states of human participants indirectly inferred from those observed social signals. Then, a pattern classifier is trained using the annotation data, and tested. In our experiments on human-robot interaction, 7 video recordings were collected and annotated with 20 social signals and 7 internal states. Several experiments were performed to obtain an 84.83% recall rate for interaction engagement, 93% for concentration intention, and 81% for task comprehension level using a C4.5 based decision tree classifier.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.6
• /
• pp.89-97
• /
• 2011

When examining current process of object tracking and search, objects were tracked by extracting them from image that was inputted through fixed single camera and objects were recognized through Zoom function to know detailed information on objects tracked. This study proposed system that expresses information on area that can seek and recognize object tracked as augmented reality by recognizing and seeking object by using multi camera. The result of experiment on proposed system showed that the number of pixels that was included in calculation was remarkably reduced and recognition rate of object was enhanced and time that took to identify information was shortened. Compared with existing methods, this system has advantage of better accuracy that can detect the motion of object and advantage of shortening time that took to detect motion.

• Animal cells and systems
• /
• v.9 no.1
• /
• pp.13-18
• /
• 2005

Cannibalism plays important roles at the levels of both individual and population. To enhance overall rate of successful survival and reproduction, salamander larvae may have evolved to consume both conspecifics and heterospecifics. Consuming conspecifics could result in decreased inclusive fitness possibly by killing relatives. In several salamander species, discrimination of salamander larval siblings from non-siblings and heterospecifics to avoid such a risk has been reported. To determine whether the Korean salamander larvae consume non-siblings more often than siblings and to analyze prey preferences of the salamander larvae in several different experimental conditions, a series of foraging experiments was conducted in the laboratory. We found that 1) large cannibal larvae preyed on small sibling more often than small non-sibling in a mixed group of sibling and non-sibling, 2) cannibal larvae prefered to consume live, weak, and small larvae to dead, healthy, and large larvae, and 3) cannibal larvae consumed heterospecific tadpoles more often than conspecific nonsibling larvae in a mixed group. In addition, the larval density was positively correlated with the occurrence of spacing behavior, one of the agonistic predator behaviors among salamander larvae.

• Bulletin of the Korean Chemical Society
• /
• v.25 no.3
• /
• pp.357-360
• /
• 2004

We developed in the present study molecular imprinted polymers (MIPs), using single templates (pentoxifylline, caffeine and theophylline) and mixed-templates (pentoxifylline-caffeine, pentoxifylline-theophylline and caffeine-theophylline). The MIPs were prepared with methacrylic acid (MAA) as the monomer, ethylene glycol dimetharylate (EGDMA) as the crosslinker and 2,2'-azobis(isobutyronitrile) (AIBN) as the initiator. The obtained polymer particles (particle size after grinding was about 25-35 ${\mu}$m) were packed into a HPLC column (3.9 mm i.d. ${\times}$ 150 mm). The selectivity and chromatographic characteristics of the MIPs were studied using acetonitrile as the mobile phase at a flow rate of 0.8 mL/min. UV detector wavelength was set at 270 nm. Different single template MIPs showed different molecular recognitions to the templates and the structurally analogues, according to the rigidity and steric hindrance of the compounds. Recognition was improved on the mixed-template MIPs as a result of the cooperation or sum effect of the templates, whereas on the pentoxifylline-theophylline imprinted polymer, the highest selectivity and affinity were obtained. Separations of the test compounds on different polymers were also investigated.

• The Journal of Korea Robotics Society
• /
• v.9 no.1
• /
• pp.1-10
• /
• 2014

In recent years, the research of 3D mapping technique in urban environments obtained by mobile robots equipped with multiple sensors for recognizing the robot's surroundings is being studied actively. However, the map generated by simple integration of multiple sensors data only gives spatial information to robots. To get a semantic knowledge to help an autonomous mobile robot from the map, the robot has to convert low-level map representations to higher-level ones containing semantic knowledge of a scene. Given a 3D point cloud of an urban scene, this research proposes a method to recognize the objects effectively using 3D graph model for autonomous mobile robots. The proposed method is decomposed into three steps: sequential range data acquisition, normal vector estimation and incremental graph-based segmentation. This method guarantees the both real-time performance and accuracy of recognizing the objects in real urban environments. Also, it can provide plentiful data for classifying the objects. To evaluate a performance of proposed method, computation time and recognition rate of objects are analyzed. Experimental results show that the proposed method has efficiently in understanding the semantic knowledge of an urban environment.

• Journal of Biomedical Engineering Research
• /
• v.30 no.3
• /
• pp.191-198
• /
• 2009

Phlegm smear testing for acid-fast bacilli (AFB) requires careful examination of tubercle bacilli under a microscope to distinguish between positive and negative findings. The biggest weakness of this method is the visual limitations of the examiners. It is also time-consuming, and mistakes may easily occur. This paper proposes a method of identifying tubercle bacilli that uses a computer instead of a human. To address the challenges of AFB testing, this study designs and investigates image systems that can be used to identify tubercle bacilli. The proposed system uses an electronic microscope to capture digital images that are then processed through feature extraction, image segmentation, image recognition, and neural networks to analyze tubercle bacilli. The proposed system can detect the amount of tubercle bacilli and find their locations. This paper analyzes 184 tubercle bacilli images. Fifty images are used to train the artificial neural network, and the rest are used for testing. The proposed system has a 95.6% successful identification rate, and only takes 0.8 seconds to identify an image.

• ETRI Journal
• /
• v.32 no.3
• /
• pp.395-405
• /
• 2010

A novel multiclassifier system (MCS) strategy is proposed and applied to a text-dependent speaker verification task. The presented scheme optimizes the linear combination of classifiers on an on-line basis. In contrast to ordinary MCS approaches, neither a priori distributions nor pre-tuned parameters are required. The idea is to improve the most accurate classifier by making use of the incremental information provided by the second classifier. The on-line multiclassifier optimization approach is applicable to any pattern recognition problem. The proposed method needs neither a priori distributions nor pre-estimated weights, and does not make use of any consideration about training/testing matching conditions. Results with Yoho database show that the presented approach can lead to reductions in equal error rate as high as 28%, when compared with the most accurate classifier, and 11% against a standard method for the optimization of linear combination of classifiers.

• Electronics and Telecommunications Trends
• /
• v.33 no.5
• /
• pp.121-134
• /
• 2018

The Von Neumann based architecture of the modern computer has dominated the computing industry for the past 50 years, sparking the digital revolution and propelling us into today's information age. Recent research focus and market trends have shown significant effort toward the advancement and application of artificial intelligence technologies. Although artificial intelligence has been studied for decades since the Turing machine was first introduced, the field has recently emerged into the spotlight thanks to remarkable milestones such as AlexNet-CNN and Alpha-Go, whose neural-network based deep learning methods have achieved a ground-breaking performance superior to existing recognition, classification, and decision algorithms. Unprecedented results in a wide variety of applications (drones, autonomous driving, robots, stock markets, computer vision, voice, and so on) have signaled the beginning of a golden age for artificial intelligence after 40 years of relative dormancy. Algorithmic research continues to progress at a breath-taking pace as evidenced by the rate of new neural networks being announced. However, traditional Von Neumann based architectures have proven to be inadequate in terms of computation power, and inherently inefficient in their processing of vastly parallel computations, which is a characteristic of deep neural networks. Consequently, global conglomerates such as Intel, Huawei, and Google, as well as large domestic corporations and fabless companies are developing dedicated semiconductor chips customized for artificial intelligence computations. The AI Processor Research Laboratory at ETRI is focusing on the research and development of super low-power AI processor chips. In this article, we present the current trends in computation platform, parallel processing, AI processor, and super-threaded AI processor research being conducted at ETRI.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.3
• /
• pp.242-247
• /
• 2008

As an absolute positioning system, iGS is designed based on ultrasonic signals whose speed can be formulated clearly in terms of time and room temperature, which is utilized for a mobile robot localization. The iGS is composed of an RFID receiver and an ultra-sonic transmitter, where an RFID is designated to synchronize the transmitter and receiver of the ultrasonic signal. The traveling time of the ultrasonic signal has been used to calculate the distance between the iGS system and a beacon which is located at a pre-determined location. This paper suggests an effective operation method of iGS to estimate position of the mobile robot working in unstructured environment. To expand recognition range and to improve accuracy of the system, two strategies are proposed: utilization of beacons belonging to neighboring blocks and removal of the environment-reflected ultrasonic signals. As the results, the ubiquitous localization system based on iGS as a pseudo-satellite system has been developed successfully with a low cost, a high update rate, and relatively high precision.