• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.891-902
• /
• 2020

The Re-Identification(Re-ID) is one of the most popular researches in the field of computer vision due to a variety of applications. To achieve a high-level re-identification performance, recently other methods have developed the deep learning based networks that are specialized for only person or vehicle. However, most of the current methods are difficult to be used in real-world applications that require re-identification of both person and vehicle at the same time. To overcome this limitation, this paper proposes a deep neural network learning method that combines triplet and softmax loss to improve performance and re-identify people and vehicles simultaneously. It's possible to learn the detailed difference between the identities(IDs) by combining the softmax loss with the triplet loss. In addition, weights are devised to avoid bias in one-side loss when combining. We used Market-1501 and DukeMTMC-reID datasets, which are frequently used to evaluate person re-identification experiments. Moreover, the vehicle re-identification experiment was evaluated by using VeRi-776 and VehicleID datasets. Since the proposed method does not designed for a neural network specialized for a specific object, it can re-identify simultaneously both person and vehicle. To demonstrate this, an experiment was performed by using a person and vehicle re-identification dataset together.

• Journal of Ocean Engineering and Technology
• /
• v.24 no.1
• /
• pp.123-132
• /
• 2010

This paper presents a prestress-loss monitoring technique for prestressed concrete (PSC) girder structures that uses a vibration-based system identification method. First, the theoretical backgrounds of the prestress-loss monitoring technique and the system identification technique are presented. Second, vibration tests are performed on a lab-scaled PSC girder for which the modal parameter was measured for several prestress-force cases. A numerical modal analysis is performed by using an initial finite element (FE) model from the geometric, material, and boundary conditions of the lab-scaled PSC girder. Third, a vibration-based system identification is performed to update the FE model by identifying structural parameters since the natural frequency of the FE model became identical to the experimental results. Finally, the feasibility of the prestress-loss monitoring technique is evaluated for the PSC girder model by using the experimentally measured natural frequency and numerically identified natural frequency for several prestress-force cases.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.472-474
• /
• 2019

Because of the difference in network structure and loss function, Verification and identification models have their respective advantages and limitations for person reidentification (re-ID). In this work, we propose a multi-task network simultaneously computes the identification loss and verification loss for person reidentification. Given a pair of images as network input, the multi-task network simultaneously outputs the identities of the two images and whether the images belong to the same identity. In experiments, we analyze the major factors affect the accuracy of person reidentification. To address the occlusion problem and improve the generalization ability of reID models, we use the Random Erasing Augmentation (REA) method to preprocess the images. The method can be easily applied to different pre-trained networks, such as ResNet and VGG. The experimental results on the Market1501 datasets show significant and consistent improvements over the state-of-the-art methods.

• Genomics & Informatics
• /
• v.16 no.4
• /
• pp.20.1-20.20
• /
• 2018

Genetic hearing loss crosses almost all the categories of hearing loss which includes the following: conductive, sensory, and neural; syndromic and nonsyndromic; congenital, progressive, and adult onset; high-frequency, low-frequency, or mixed frequency; mild or profound; and recessive, dominant, or sex-linked. Genes play a role in almost half of all cases of hearing loss but effective treatment options are very limited. Genetic hearing loss is considered to be extremely genetically heterogeneous. The advancements in genomics have been instrumental to the identification of more than 6,000 causative variants in more than 150 genes causing hearing loss. Identification of genes for hearing impairment provides an increased insight into the normal development and function of cells in the auditory system. These defective genes will ultimately be important therapeutic targets. However, the auditory system is extremely complex which requires tremendous advances in gene therapy including gene vectors, routes of administration, and therapeutic approaches. This review summarizes and discusses recent advances in elucidating the genomics of genetic hearing loss and technologies aimed at developing a gene therapy that may become a treatment option for in the near future.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.5
• /
• pp.1755-1777
• /
• 2022

The development of wireless communication technology has led to the underutilization of radio spectra. To address this limitation, an intelligent cognitive radio network was developed. Specific emitter identification (SEI) is a key technology in this network. However, in realistic non-cooperative scenarios, the system may detect signal classes beyond those in the training database, and only a few labeled signal samples are available for network training, both of which deteriorate identification performance. To overcome these challenges, a meta-learning-based open-set identification system is proposed for SEI. First, the received signals were pre-processed using bi-spectral analysis and a Radon transform to obtain signal representation vectors, which were then fed into an open-set SEI network. This network consisted of a deep feature extractor and an intrinsic feature memorizer that can detect signals of unknown classes and classify signals of different known classes. The training loss functions and the procedures of the open-set SEI network were then designed for parameter optimization. Considering the few-shot problems of open-set SEI, meta-training loss functions and meta-training procedures that require only a few labeled signal samples were further developed for open-set SEI network training. The experimental results demonstrate that this approach outperforms other state-of-the-art SEI methods in open-set scenarios. In addition, excellent open-set SEI performance was achieved using at least 50 training signal samples, and effective operation in low signal-to-noise ratio (SNR) environments was demonstrated.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.55 no.4
• /
• pp.294-302
• /
• 2019

As a series of fundamental researches on the development of an automatic identification monitoring system for fishing gear. Firstly, the study on the installation method of automated identification buoy for the coastal improvement net fishing net with many loss problems on the west coast was carried out. Secondly, the study was conducted find out how to install an automatic identification buoy for coastal gill net which has the highest loss rate among the fisheries. GPS for fishing was used six times in the coastal waters around Seogwipo city in Jeju Island to determine the developmental status and underwater behavior to conduct a field survey. Next, a questionnaire was administered in parallel on the type of loss and the quantity and location of fishing gear to be developed and the water transmitter. In the field experiment, the data collection was possible from a minimum of 13 hours, ten minutes to a maximum of 20 hours and ten minutes using GPS, identifying the development status and underwater behavior of the coastal gillnet fishing gear. The result of the survey showed that the loss of coastal net fishing gear was in the following order: net (27.3%), full fishing gear (24.2%), buoys, and anchors (18.2%). The causes were active algae (50.0%), fish catches (33.3%) and natural disasters (12.5%). To solve this problem, the installation method is to attach one and two electronic buoys to top of each end of the fishing gear, and one underwater transmitter at both ends of the float line connected to the anchor. By identifying and managing abnormal conditions such as damage or loss of fishing gear due to external factors such as potent algae and cutting of fishing gear, loss of fishing gear can be reduced. If the lost fishing gear is found, it will be efficiently collected.

• Structural Engineering and Mechanics
• /
• v.17 no.3_4
• /
• pp.467-482
• /
• 2004

One of the uncertain damage parameters to jeopardize the safety of existing PSC bridges is the loss of the prestress force. A substantial prestress-loss can lead to severe problems in the serviceability and safety of the PSC bridges. In this paper, a nondestructive method to detect prestress-loss in beam-type PSC bridges using a few natural frequencies is presented. An analytical model is formulated to estimate changes in natural frequencies of the PSC bridges under various prestress forces. Also, an inverse-solution algorithm is proposed to detect the prestress-loss by measuring the changes in natural frequencies. The feasibility of the proposed approach is evaluated using PSC beams for which a few natural frequencies were experimentally measured for a set of prestress-loss cases. Numerical models of two-span continuous PSC beams are also examined to verify that the proposed algorithm works on more complicated cases.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.5
• /
• pp.415-422
• /
• 2013

This paper proposes a robust binary audio fingerprinting method by minimizing hinge-loss function. In the proposed method, the type of fingerprints is binary, which is conducive in reducing the size of fingerprint DB. In general, the binarization of features for fingerprinting deteriorates the performance of fingerprinting system, such as robustness and discriminability. Thus it is necessary to minimize such performance loss. Since the similarity between two audio clips is represented by a hinge-like function, we propose a method to derive a binary fingerprinting by minimizing a hinge-loss function. The derived hinge-loss function is minimized by using the minimal loss hashing. Experiments over thousands of songs demonstrate that the identification performance of binary fingerprinting can be improved by minimizing the proposed hinge loss function.

• Journal of the Korea Safety Management & Science
• /
• v.6 no.2
• /
• pp.155-165
• /
• 2004

It is needed to develop on-line real time management and RFID-PPS(Radio Frequency Identification-Pallet Pool System) by putting information technology. Additionally, it is possible to figure out the flow of all the materials loaded on the RFID pallet; product, material, raw material immediately, so that epoch-making management is possible and it contributes to the reduction of logistics cost because there are little loss or outflow of pallet. The materials flow is getting speedy and inventory is decreasing in the logistics process, and also bad inventory and loss problems are prevented. As a result, not only logistics cost of company but also national logistics cost is decreased. Thus it contributes to the strength of national competitiveness.

• Journal of the Korean Society for Railway
• /
• v.6 no.2
• /
• pp.9-18
• /
• 2003

This paper discuss advanced safety in the railway signaling system. The specified methods and HAZOP about Hazard identification and analysis of railway signalling system were studied, and loss analysis and ALARP model in order to calculate safety as a standard capacity were proposed. It was also resulted from Hazard identification, analysis and evaluation by applying advanced safety to the railway signalling system.