• Journal of Convergence for Information Technology
• /
• v.11 no.6
• /
• pp.33-39
• /
• 2021

Speech recognition technology is being combined with deep learning and is developing at a rapid pace. In particular, voice recognition services are connected to various devices such as artificial intelligence speakers, vehicle voice recognition, and smartphones, and voice recognition technology is being used in various places, not in specific areas of the industry. In this situation, research to meet high expectations for the technology is also being actively conducted. Among them, in the field of natural language processing (NLP), there is a need for research in the field of removing ambient noise or unnecessary voice signals that have a great influence on the speech recognition recognition rate. Many domestic and foreign companies are already using the latest AI technology for such research. Among them, research using a convolutional neural network algorithm (CNN) is being actively conducted. The purpose of this study is to determine the non-voice section from the user's speech section through the convolutional neural network. It collects the voice files (wav) of 5 speakers to generate learning data, and utilizes the convolutional neural network to determine the speech section and the non-voice section. A classification model for discriminating speech sections was created. Afterwards, an experiment was conducted to detect the non-speech section through the generated model, and as a result, an accuracy of 94% was obtained.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.3
• /
• pp.59-66
• /
• 2021

Humans mainly recognize surrounding objects using visual and auditory information among the five senses (sight, hearing, smell, touch, taste). Major research related to the latest object recognition mainly focuses on analysis using image sensor information. In this paper, after emitting various chirp audio signals into the observation space, collecting echoes through a 2-channel receiving sensor, converting them into spectral images, an object recognition experiment in 3D space was conducted using an image learning algorithm based on deep learning. Through this experiment, the experiment was conducted in a situation where there is noise and echo generated in a general indoor environment, not in the ideal condition of an anechoic room, and the object recognition through echo was able to estimate the position of the object with 83% accuracy. In addition, it was possible to obtain visual information through sound through learning of 3D sound by mapping the inference result to the observation space and the 3D sound spatial signal and outputting it as sound. This means that the use of various echo information along with image information is required for object recognition research, and it is thought that this technology can be used for augmented reality through 3D sound.

• Journal of Broadcast Engineering
• /
• v.27 no.2
• /
• pp.244-247
• /
• 2022

In this paper, the performance of the decoding schemes using linear MMSE filters in the frequency and time domains and the reinforcement Gauss-Seidel algorithm for the orthogonal time frequency space (OTFS) system that can improve robustness under high-speed mobile environments are compared. The reinforcement Gauss-Seidel algorithm can improve the bit error rate performance by suppressing the noise enhancement. The simulation results show that the performance of the decoding scheme using the linear MMSE filter in the frequency domain is severely degraded due to the effect of Doppler shift as the mobile speed increases. In addition, the decoding scheme using the reinforcement Gauss-Seidel algorithm under the channel environment with 120 km/h and 500 km/h speeds outperforms the decoding schemes using linear MMSE filters in the frequency and time domains.

• Analytical Science and Technology
• /
• v.19 no.2
• /
• pp.142-148
• /
• 2006

To determine malachite green and leuco-malachite green residues in sea food, a liquid chromatographic method has been optimized. The target compounds were extracted in the homogenized edible tissues with a mixture of McIlvaine buffer-acetonitrile and partitioned against dichloromethane. After concentrating the lower layer, the resulting residues were re-dissolved in methanol and analyzed by the HPLC with visible detector at 620 nm using acetonitrile-acetate buffer. For the analysis of leuco-malachite green with malachite green simultaneously, post-column packed with lead(IV) oxide was used for oxidizing leuco-malachite green to malachite green. The correlation coefficients($r^2$) was 0.9989 for malachite green, and 0.9995 for leuco-malachite green. The limit of detection was 0.005 mg/kg for the combined of malachite green and leuco-malachite green at signal/noise${\geq}3$. The recovery rate was within a reliable range of 84~98% (CV 3~16%). Leuco-malachite green were detected in carp and crusian carp.

• Analytical Science and Technology
• /
• v.20 no.1
• /
• pp.84-90
• /
• 2007

To determine levels of 11 sulfonamides for animals in food, simultaneously, a selective method of high performance liquid chromatography with UV detector has been applied. The targets were sulfachlorpyridazine (SCP), sulfadiazine (SDZ), sulfadimethoxine (SDM), sulfisoxazole (SSX), sulfamerazine (SMZ), sulfamethazine (SMT), sulfamethoxazole (SMX), sulfamethoxypyridazine (SMP), sulfamonomethoxine (SMM), sulfaquinoxaline (SQX) and sulfathiazole (STZ). Food samples were beef, pork, chicken, milk and whole egg that were collected at the main 6 cities in Korea as Seoul, Busan, Daejon, Incheon, Mokpo and Gangneung. After homogenizing food samples with sodium phosphate solution and acetonitrile, it was extracted with n-hexane. The mobile phase gradient was a mixture of 5 mM potassium phosphate (pH 3.25) and methanol with a gradient ratio from 100:0 to 30:70. The UV wavelength was 270 nm. The overall recoveries were ranged from 75% to 95% and the limit of detection was minimum 0.004 mg/kg for SMT, and 0.007 mg/kg for STZ at signal/noise > 3, respectively. As results, sulfonamide drugs were not detected in most of the selected food samples, however, sulfamonomethoxine was detected in meat. The determined level of sulfamonomethoxine were 0.03 and 0.06 mg/kg for beef that were below the MRLs.

• Korean Journal of Radiology
• /
• v.25 no.9
• /
• pp.798-806
• /
• 2024

Objective: Dual-layer CT (DLCT) can create virtual monochromatic images (VMIs) at various monochromatic X-ray energies, particularly at low keV levels, with high contrast-to-noise ratio. The purpose of this study was to assess the clinical feasibility of contrast-enhanced chest DLCT with a low keV VMI for preoperative breast cancer staging, in comparison to breast MRI. Materials and Methods: A total of 152 patients with 155 index breast cancers were enrolled in the study. VMIs were generated from contrast-enhanced chest DLCT at 40 keV and maximum intensity projection (MIP) with three-dimensional (3D) reconstruction was performed for both bilateral breast areas. Two radiologists reviewed in consensus the 3D MIP images of the chest DLCT with VMI and breast MRI in separate sessions with a 3-month wash-out period. The detection rate and mean tumor size of the index cancer were compared between the chest DLCT with VMI and breast MRI. Additionally, the agreement of tumor size measurement between the two imaging modalities were evaluated. Results: Of all index cancers, 84.5% (131/155) were detected in the chest DLCT with VMI, while 88.4% (137/155) were detected in the breast MRI (P = 0.210). The Bland-Altman agreement between the chest DLCT with VMI and breast MRI was a mean difference of -0.05 cm with 95% limits of agreement of -1.29 to 1.19 cm. The tumor size in the chest DLCT with VMI (2.3 ± 1.7 cm) was not significantly different from that in the breast MRI (2.4 ± 1.6 cm) (P = 0.106). Conclusion: The feasibility of chest DLCT with VMI was demonstrated for preoperative tumor staging in breast cancer patients, showing comparable cancer detectability and good agreement in tumor size measurement compared to breast MRI. This suggests that chest DLCT with VMI can serve as a potential alternative for patients who have contraindications to breast MRI.

• Journal of Animal Science and Technology
• /
• v.49 no.5
• /
• pp.559-568
• /
• 2007

Recently, copy number variations (CNV) of genes or genomic segments have been intensively studied and various analysis methods have been developed. In this study, quantitative oligonucleotide ligation assay (qOLA) was applied to investigate CNV of KIT gene in the Landrace breed. A combined assay using qOLA and pyrosequencing, 6 genotype classes, I1/I1 or I3/i (IBe), I1/I2 or I3/IP, I1/I3, I1/IP or I2/i (IBe), I2/I2and I2/IP, were identified from 44 Landrace pigs. Genotype assignment using grouping features of measurements on a scatter plot showed 100% agreement with those using a statistical assignment by PROC FASTCLUS procedure implemented in the SAS package. Two versions (3100 and 3130) of ABI sequencers gave the same genotyping results, indicating there was no influence on qOLA by different versions of instrument, however, the means of standard deviation and coefficient of variation from the qOLA on a ABI 3130 (2.33 and 4.10) was lower than those from the qOLA on a ABI 3100 (2.67 and 4.81). Effect of proteinase K treatment on the PCR product followed by qOLA was very clear because noise peaks were disappeared and the observed ration fit better to the reference ratio corresponding to each genotype.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2004.04a
• /
• pp.5-9
• /
• 2004

In this paper, we studied the OCT(Optical Coherence Tomography) system which it has been extensively studied because of having some advantages such as high resolution cross-sectional images, low cost, and small size configuration. A basic principle of OCT system is Michelson interferometer. The characteristics of light source determine the resolution and the transmission depth. As a results, the light source have a commercial SLD with a central wavelength of 1,285 nm and FWHM(Full Width at Half Maximum) of 35.3 nm. The optical delay line part is necessary to equal of the optical path length with scattered light or reflected light from sample. In order to equal the optical path length, the stage which is attached to reference mirror is moved linearly by step motor And the interferometer is configured with the Michelson interferometer using single mod fiber, the scanner can be focused of the sample by using the reference arm. Also, the 2-dimensional cross-sectional images were measured with scanning the transverse direction of the sample by using step motor. After detecting the internal signal of lateral direction at a paint of sample, scanner is moved to obtain the cross-sectional image of 2-demensional by using step motor. Photodiode has been used which has high detection sensitivity, excellent noise characteristic, and dynamic range from 800 nm to 1,700 nm. It is detected mixed small signal between noise and interference signal with high frequency After filtering and amplifying this signal, only envelope curve of interference signal is detected. And then, cross-sectional image is shown through converting this signal into digitalized signal using A／D converter. The resolution of the OCT system is about 30$\mu\textrm{m}$ which corresponds to the theoretical resolution. Also, the cross-sectional image of ping-pong ball is measured. The OCT system is configured with Michelson interferometer which has a low contrast because of reducing the power of feedback interference light. Such a problem is overcomed by using the improved inteferometer. Also, in order to obtain the cross-sectional image within a short time, it is necessary to reduce the measurement time for improving the optical delay line.

• Proceedings of the KOR-BRONCHOESO Conference
• /
• 1978.06a
• /
• pp.9.2-9
• /
• 1978

Monaural and binaural hearing aid performance under quiet and noisy conditions were compared in regard to (1) the degree of hearing impairment, (2) the symmetry of pure tone audiogram, (3) the automatic gain control of the hearing aid. (4) hearing impairement with recruitment and, word discrimination ability. Performance using binaural hearing aids was consistently superior to that using monaural hearing aids. The results were as follows. 1. Speech detection thresholds were enhanced by a mean of 4.25dB when tested with danavox 747 PP stereo type hearing aid and by a mean of 4.12 dB when tested hearing aids connected seperately to the right and left ears. 2. Binaurally tested speech reception thresholds were superior to monaurally tested thresholds by a mean of 3.56dB when tested in quiet and by a mean of 5.56dB when tested in noise. 3. Binaurally tested word discrimination scores were also superior by a mean of 17.09% in quiet and by a mean 19.63% in noise. 4. Both SRT and word discrimination scores were performed best by subjects with moderately-severe impairement. The performance by one mildly impaired subject was the poorest of all performances. The levels of performance order were; moderately-severe loss, severe loss. moderate loss and mild loss. 5. The data obtained using AGC aids when compaired with that of linear amplification show that when AGC aids were worn in both ears. the results were very poor but when one AGC aid was worn in one ear and linear amplification in the other. the results were good. 6. The advantages of binaural hearing aids were obvious even in cases 1) with great diferences in hearing thresholds between right and left ears, 2) when the subject was unable to discriminate words without vision and. 3) when the subject had extreme recruitme t phenomenon.

• Journal of the Korean Society of Radiology
• /
• v.82 no.6
• /
• pp.1477-1492
• /
• 2021

Purpose Dedicated breast CT is an emerging volumetric X-ray imaging modality for diagnosis that does not require any painful breast compression. To improve the detection rate of weakly enhanced lesions, an adaptive image rescaling (AIR) technique was proposed. Materials and Methods Two disks containing five identical holes and five holes of different diameters were scanned using 60/100 kVp to obtain single-energy CT (SECT), dual-energy CT (DECT), and AIR images. A piece of pork was also scanned as a subclinical trial. The image quality was evaluated using image contrast and contrast-to-noise ratio (CNR). The difference of imaging performances was confirmed using student's t test. Results Total mean image contrast of AIR (0.70) reached 74.5% of that of DECT (0.94) and was higher than that of SECT (0.22) by 318.2%. Total mean CNR of AIR (5.08) was 35.5% of that of SECT (14.30) and was higher than that of DECT (2.28) by 222.8%. A similar trend was observed in the subclinical study. Conclusion The results demonstrated superior image contrast of AIR over SECT, and its higher overall image quality compared to DECT with half the exposure. Therefore, AIR seems to have the potential to improve the detectability of lesions with dedicated breast CT.