• Journal of Internet Computing and Services
• /
• v.22 no.6
• /
• pp.33-40
• /
• 2021

Automatic Speech Recognition(ASR) is a technology that analyzes human speech sound into speech signals and then automatically converts them into character strings that can be understandable by human. Speech recognition technology has evolved from the basic level of recognizing a single word to the advanced level of recognizing sentences consisting of multiple words. In real-time voice conversation, the high recognition rate improves the convenience of natural information delivery and expands the scope of voice-based applications. On the other hand, with the active application of speech recognition technology, concerns about related cyber attacks and threats are also increasing. According to the existing studies, researches on the technology development itself, such as the design of the Automatic Speaker Verification(ASV) technique and improvement of accuracy, are being actively conducted. However, there are not many analysis studies of attacks and threats in depth and variety. In this study, we propose a cyber attack model that bypasses voice authentication by simply manipulating voice frequency and voice speed for AI voice recognition service equipped with automated identification technology and analyze cyber threats by conducting extensive experiments on the automated identification system of commercial smartphones. Through this, we intend to inform the seriousness of the related cyber threats and raise interests in research on effective countermeasures.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.10
• /
• pp.55-61
• /
• 2009

This paper describes a point of production(POP) system which collects and manages real-time shop floor machining information in a wafer dicing process. The system are composed of POP terminal, line controller and network. In the configuration of the system, LAN and RS485 network are used for connection with the upper management system and down stratum respectively. As a bridge between POP terminal and server, a line controller is used. The real-time information which is the base of production management are collected from information resources such as machine, product and worker. The collected information are used for the calculation of optimal cutting condition. The collection of the information includes cutting speed, spout of pure water, accumulated count of cut in process for blade and wafer defect. In order to manage machining information in wafer dicing process, production planning information is delivered to the shop floor, and production result information is collected from the shop floor, delivered to the server and used for managing production plan. From the result of the system application, production progress status, work and non-working hour analysis for each machine, and wafer defect analysis are available, and they are used for quality and productivity improvements in wafer dicing process. A case study is implemented to evaluate the performance of the system.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.9
• /
• pp.95-104
• /
• 2009

This paper proposes the noise robust algorithm to detect the starting point of music. Detection of starting point of music is necessary to solve computational-waste problem and retrieval-comparison problem with inconsistent input data in music content based retrieval system. In particular, such detection is even more necessary in time sequential retrieval method that compares data in the sequential order of time in contents based music retrieval system. Whereas it has the long point that the retrieval is fast since it executes simple comparison in the order of time, time sequential retrieval method has the short point that data starting time to be compared should be the same. However, digitalized music cannot guarantee the equity of starting time by bit rate conversion. Therefore, this paper ensured that recognition rate shall not decrease even while executing high speed retrieval by applying time sequential retrieval method through detection of music starting point in the pre-processing stage of retrieval. Starting point detection used minimum wave model that can detect effective sound, and for strength against noise, the noises existing in mute sound were swapped. The proposed algorithm was confirmed to produce about 38% more excellent performance than the results to which starting point detection was not applied, and was verified for the strength against noise.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.34 no.1
• /
• pp.1-7
• /
• 2024

β-Ga₂O₃ is a material with a wide band gap of ~4.8 eV and a high breakdown-voltage of 8 MV/cm, and is attracting much attention in the field of power device applications. In addition, compared to representative WBG semiconductor materials such as SiC, GaN and Diamond, it has the advantage of enabling single crystal growth with high growth rate and low manufacturing cost [1-4]. In this study, we succeeded in growing a 10 mm thick β-Ga₂O₃ single crystal doped with 0.3 mol% SnO₂ through the EFG (Edge-defined Film-fed Growth) method using multi-slit structure. The growth direction and growth plane were set to [010]/(010), respectively, and the growth speed was about 12 mm/h. The grown β-Ga₂O₃ single crystal was cut into various crystal planes (010, 001, 100, ${\bar{2}}01$) and surface processed. The processed samples were compared for characteristics according to crystal plane through analysis such as XRD, UV/VIS/NIR/Spec., Mercury Probe, AFM and Etching. This research is expected to contribute to the development of power semiconductor technology in high-voltage and high-temperature applications, and selecting a substrate with better characteristics will play an important role in improving device performance and reliability.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.44 no.1
• /
• pp.109-116
• /
• 2024

The Rapid Hardening Composite Mat (RHCM) is a product that improves the initial strength development speed of conventional Geosynthetic Cementitious Composite Mats (GCCM). It offers the advantage of quickly securing sufficient strength in railway slopes with insufficient formation level, and provides benefits such as preventing slope erosion and inhibiting vegetation growth. In this study, an analysis of the practical applicability of RHCM in railway settings was conducted through experimentation. The on-site applicability was assessed by categorizing it into fire resistance, durability, and stability, and conducting combustibility test, ground contact pressure test, and daily displacement analyses. In the case of South Korea, where a significant portion of the territory is composed of forested areas, the prevention of slope fires is imperative. To analyze the fire resistance of RHCM, combustibility tests were conducted as an essential measure. Durability was assessed through ground contact pressure tests to analyze the deformation and potential damage of RHCM caused by the inevitable use of small to medium-sized equipment on the construction surface. Furthermore, daily displacement analysis was conducted to evaluate the structural stability by comparing and analyzing the displacement and behavior occurring during the application of RHCM with railway slope maintenance criteria. As a result of the experiments, the RHCM was analyzed to meet the criteria for heat release rate and gas toxicity. Furthermore, the ground contact pressure was observed to be consistently above 50 kPa during the curing period of 4 to 24 hours under all conditions. Additionally, the daily displacement analyzed through field site experiments ranged from -1.7 mm to 1.01 mm, confirming compliance with the criteria.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.6
• /
• pp.250-258
• /
• 2007

Current speech recognition technology s achieved high performance with the development of hardware devices, however it is insufficient for some applications where high reliability is required, such as voice control of powered wheelchairs for disabled persons. For the system which aims to operate powered wheelchairs safely by voice in real environment, we need to consider that non-voice commands such as user s coughing, breathing, and spark-like mechanical noise should be rejected and the wheelchair system need to recognize the speech commands affected by disability, which contains specific pronunciation speed and frequency. In this paper, we propose non-voice rejection method to perform voice/non-voice classification using both YIN based fundamental frequency(F0) extraction and reliability in preprocessing. We adopted a multi-template dictionary and acoustic modeling based speaker adaptation to cope with the pronunciation variation of inarticulately uttered speech. From the recognition tests conducted with the data collected in real environment, proposed YIN based fundamental extraction showed recall-precision rate of 95.1% better than that of 62% by cepstrum based method. Recognition test by a new system applied with multi-template dictionary and MAP adaptation also showed much higher accuracy of 99.5% than that of 78.6% by baseline system.

• Tuberculosis and Respiratory Diseases
• /
• v.48 no.4
• /
• pp.448-463
• /
• 2000

Background : Pulmonary infiltrate is a frequent cause of morbidity and mortality in patients with leukemia. It is often hard to obtain a reliable diagnosis by clinical and radiologic findings alone. The aim of this study was to evaluate diagnostic and therapeutic benefits of invasive procedures for new lung infiltrates in leukemia. Methods : Patients with leukemia who developed new lung infiltrates from December 1994 to March 1999 were included in this study. These patients were classified into the empirical group who received empirical therapy only and into the invasive group who underwent bronchoscopy or surgical lung biopsy for the diagnostic purpose of new lung infiltrates. A retrospective chart review was done to find the etiologies of new lung infiltrates, the yield of invasive procedures, outcome as well as predicting factors for survival. Results : 1) One hundred-two episodes of new lung infiltrates developed in 90 patients with leukemia. Invasive procedures were performed in 44 episodes while 58 episodes were treated with empirical therapy only. 2) Invasive procedures yielded a specific diagnosis in 72.7%(32/44), of which 78.1% had infectious etiology. Therapeutic plan was changed in 52.3%(23/44) of patients after invasive procedures. None of them showed procedure-related mortality. 3) The overall survival rate was 62.7%(64/102). Survival rate in the invasive group (79.5%) was significantly better than that in the empirical group (50.0%) (p=0.002). 4) Upon multivariate analysis, the performance of invasive procedures, no need for mechanical ventilation and achievement of complete remission of leukemia after induction chemotherapy were the independent predicting factors for survival in patients with leukemia and new lung infiltrates. Conclusion : Bronchoscopy and surgical lung biopsy are useful in the diagnosis of new lung infiltrates in patients with leukemia. However, survival benefits of invasive procedures should be considered together with disease status of leukemia and severity of respiratory compromise.

• Journal of Korea Society of Industrial Information Systems
• /
• v.12 no.4
• /
• pp.15-24
• /
• 2007

The recognition of postal address images is indispensable for the automatic sorting of postal envelopes. The process of the address image recognition is composed of three steps-address image preprocessing, character recognition, address interpretation. The extracted character images from the preprocessing step are forwarded to the character recognition step, in which multiple candidate characters with reliability scores are obtained for each character image extracted. aracters with reliability scores are obtained for each character image extracted. Utilizing those character candidates with scores, we obtain the final valid address for the input envelope image through the address interpretation step. The envelope sorting rate depends on the performance of all three steps, among which character recognition step could be said to be very important. The good character recognizer would be the one which could produce valid candidates with very reliable scores to help the address interpretation step go easy. In this paper, we propose the method of generating character candidates with reliable recognition scores. We utilize the existing MLP(multilayered perceptrons) neural network of the address recognition system in the current automatic postal envelope sorters, as the classifier for the each image from the preprocessing step. The MLP is well known to be one of the best classifiers in terms of processing speed and recognition rate. The false alarm problem, however, might be occurred in recognition results, which made the address interpretation hard. To make address interpretation easy and improve the envelope sorting rate, we propose promising methods to reestimate the recognition score (confidence) of the existing MLP classifier: the generation method of the statistical recognition properties of the classifier and the method of the combination of the MLP and the subspace classifier which roles as a reestimator of the confidence. To confirm the superiority of the proposed method, we have used the character images of the real postal envelopes from the sorters in the post office. The experimental results show that the proposed method produces high reliability in terms of error and rejection for individual characters and non-characters.

• Korean Chemical Engineering Research
• /
• v.47 no.2
• /
• pp.220-229
• /
• 2009

Strain improvement and morphology investigation in bioreactor cultures were undertaken in suspended cultures of Phellinus linteus mycelia for mass production of protein-bound polysaccharides(soluble ${\beta}$-D-glucan), a powerful immuno-stimulating agent. Phellineus sp. screened for this research was identified as Phellinus linteues through ITS rDNA sequencing method and blast search, demonstrating 99.7% similarity to other Phellinus linteus strains. Intensive strain improvement program was carried out by obtaining large amounts of protoplasts for the isolation of single cell colonies. Rapid and large screening of high-yielding producers was possible because large numbers of protoplasts ($1{\times}10^5{\sim}10^6\;protoplasts/ml$) formed using the banding filtration method with the cell wall-disrupting enzymes could be regenerated in relatively high regeneration frequency($10^{-2}{\sim}10^{-3}$) in the newly developed regeneration medium. It was demonstrated that the strains showing high performances in the protoplast regeneration and solid growth medium were able to produce 5.8~6.4%(w/w) of ${\beta}$-D-glucan and 13~15 g/L of biomass in stable manners in suspended shake-flask cultures of P. linteus mycelia. In addition, cell mass increase was observed to be the most important in order to enhance ${\beta}$-D-glucan productivity during the course of strain improvement program, since the amount of ${\beta}$-D-glucan extracted from the cell wall of P. linteus mycelia was almost constant on the unit biomass basis. Therefore we fully investigated the fungal cell morphology, generally known as one of the key factors affecting cell growth extent in the bioreactor cultures of mycelial fungal cells. It was found that, in order to obtain as high cell mass as possible in the final production bioreactor cultures, the producing cells should be proliferated in condensed filamentous forms in the growth cultures, and optimum amounts of these filamentous cells should be transferred as active inoculums to the production bioreactor. In this case, ideal morphologies consisting of compacted pellets less than 0.5mm in diameter were successfully induced in the production cultures, resulting in shorter period of lag phase, 1.5 fold higher specific cell growth rate and 3.3 fold increase in the final biomass production as compared to the parallel bioreactor cultures of different morphological forms. It was concluded that not only the high-yielding but also the good morphological characteristics led to the significantly higher biomass production and ${\beta}$-D-glucan productivity in the final production cultures.

• Reproductive and Developmental Biology
• /
• v.29 no.3
• /
• pp.163-168
• /
• 2005

The aim of this study was to investigate the changes in insulin-like growth factor-I (IGF-I) and growth hormone (GH) in serum, the quantitation of spermato-genesis and the comparable relationships among these measurements during pubertal period in New Zealand White male rabbits. To investigate the age-related testicular changes in DNA contents of spermatogenic cells, the fine-needle testicular biopsies from males aged 10 to 28 wks were evaluated by flow cytometry(FCM). Body weight increased significantly between the ages of 12 and 20 wks (P<0.05) and reached 3.4 kg at 28 wks of age. The highest serum IGF-I level (451.3ng/mL) was observed at 20wks of age (P<0.05) and thereafter remained stable at low levels. Serum GH level at 18 wks of age was 183.3 pg/mL which was significantly higher compared to the other ages (P<0.05), and the rising time in serum GH tend to be somewhat earlier than that of IGF-I. The relative percentage of It-cells in testicular cell compartments was $48.2\%$ at the age of 18 wks which significantly increased than those of 16-wk-old (P<0.05) and thereafter increased with the advance of age to $68\%$. The percentage of 2C-cells in testis was $26.8\%$ at 18 wks of age which was significantly lower than $54.3\%$ at 16 wks old (P<0.05). The percentage of 4C-cells was constantly maintained $2\~6\%$ except the $9.9\%$ at 18 wks of age. In conclusion, the results suggest that the puberty onset occurred at about the 18 wks of age and that the IGF-I and GH in serum during the pubertal period showed the age/growth-specific changes and these changes might be related to the spermatogenesis. The DNA FCM combined with fine-needle testicular biopsy could offer a very sensitive method to monitor the quantitative spermatogenic events related to the puberty onset.