• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.5
• /
• pp.1-8
• /
• 2017

Recently, with the development of ICT technology such as cloud and mobile, image utilization through social networks is increasing rapidly. These images contain personal information, and personal information leakage accidents may occur. As a result, studies are underway to recognize and mask personal information in images. However, optical character recognition, which recognizes personal information in images, varies greatly depending on brightness, contrast, and distortion, and Korean recognition is insufficient. Therefore, in this paper, we design and implement a personal information identification and masking system based on image recognition through deep learning application using CNN algorithm based on optical character recognition method. Also, the proposed system and optical character recognition compares and evaluates the recognition rate of personal information on the same image and measures the face recognition rate of the proposed system. Test results show that the recognition rate of personal information in the proposed system is 32.7% higher than that of optical character recognition and the face recognition rate is 86.6%.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.22 no.4
• /
• pp.121-130
• /
• 2014

This paper is the second investigation on the effects of intake flow control methods on the part load performance in a spark ignition engine. In the previous work, two control methods, port throttling and masking, were compared with respect to lean misfire limit, fuel consumption and emissions. In this work, the effects of these two methods on EGR characteristics were studied and simultaneously the differences between EGR and lean combustion as a dilution method were investigated. The results show that EGR limit is expanded up to 23% and 3 ~ 5% improvement in the fuel consumption are achieved around 8 ~ 13% rates by the flow controls comparing with 10% limit and 1.5% reduction around 3% rate of non-control case. The masking method is more effective on the limit expansion than throttling as like as lean misfire limit; however there is no substantial difference in fuel consumptions improvement regardless the control methods except high load condition. Also it is observed that there exist critical EGR rates around which the combustion performance and NOx formation change remarkably and these rates generally coincide with optimum rates for the fuel consumption. In addition, dilution with fresh air is much more advantageous than that of the exhaust gas from the view point of dilution limit and fuel consumption, while utilization of the exhaust gas is more effective on NOx reduction in spite of considerably small dilution compared with the use of fresh air. Finally, the improvement of fuel consumption by massive EGR is highly dependent on the EGR limit at which the engine runs stably, therefore the stratified combustion technique might be a best solution for this purpose.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.48 no.4
• /
• pp.479-486
• /
• 2012

In order to obtain the fundamental data about the behavior of conger by underwater audible sound, this experiment was carried out to investigate the hearing ability of Conger eel Conger myriaster which was in the coast of Jeju Island by heartbeat conditioning method using pure tones coupled with a delayed electric shock. The audible range of conger eel extended from 50Hz to 300Hz with a peak sensitivity at 80Hz including less sensitivity over 200Hz. The mean auditory thresholds of conger eel at the frequencies of 50Hz, 80Hz, 100Hz, 200Hz and 300Hz were 105dB, 92dB, 96dB, 128dB and 140dB, respectively. The positive response of conger eel was not evident after the sound projection of over 200Hz. At the results, the sensitive frequency range of conger eel is narrow in spite of swim bladder. Auditory masking was determined for Conger eel by using masking stimuli with the spectrum level range of about 60~70dB (0dB re $1{\mu}Pa/\sqrt{Hz}$). According to white noise level, the auditory thresholds increased as compared with thresholds in a quiet background noise including critical ratio at 68dB of white noise from minimum 26dB to maximum 30dB at test frequencies of 80Hz and 100Hz. The noise spectrum level at the start of masking was distributed at the range of about 68dB within 80~100Hz.

• The Journal of Engineering Research
• /
• v.2 no.1
• /
• pp.57-62
• /
• 1997

Two psychoacoustic models used in MPEG-1 employ different masking patterns, different masking indexes, and different computational procedures. As a result, Model 1 is inferior to Model 2 due to its worst case approach in computing the SMR even though it determines tonality and masking levels accurately. In this study, we investigate the performances of psychoacoustic models when we modify the MPEG-1 audio coder for variable bitrates. Simulation results show that Model 2 has a gain of 30 kbps in the dual channel mode and 20 kbps in the joint stereo mode. It is generally known that the joint stereo mode has a gain in bitrate compare to the dual channel mode. For signals with frequent attacks, this gain becomes larger in Model 1 than in Model 2. This is due to the fact that Model 1 uses the worst case approach in computing the SMR to reduce pre-echo

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.6
• /
• pp.698-703
• /
• 2003

This paper presents analog secure communication system about safe speech transmission using chaotic signals. We applied various conditions that happen in actuality communication environment modifying chaotic synchronization and chaotic communication schemes and analyzed restoration performance of speech signal to computer simulation. In transmitter, we made the chaotic masking signal which is added voice signal to chaotic signal using PC(Pecora ＆ Carroll) and SFB(self-feedback) control techniques and transmitted encryption signal to noisy communication channel And in order to calculate the degree of restoration performance, we proposed the definition of analog average power of recovered error signals in receiver chaotic system. The simulation results show that feedback control techniques can certify that restoration performance is superior to quantitative data than PC method about masking degree, susceptibility of parameters and channel noise. We experimentally computed the table of relation of parameter fluxion to restoration error rate which is applied the encryption key values to the chaotic secure communication.

• Journal of Fisheries and Marine Sciences Education
• /
• v.28 no.2
• /
• pp.384-390
• /
• 2016

The purpose of this paper is to improve the availability of underwater sound by the fundamental data on the hearing ability of Redlip croaker Pseudosciaena polyactis, which is cultured according to the cultivation technology, recently. The auditory thresholds of Redlip croaker were determined at 6 frequencies from 80Hz to 800Hz by heartbeat conditioning method using pure tones coupled with a delayed electric shock. The audible range of the Redlip croaker extended from 80Hz to 800Hz with the best sensitive frequency range including little difference in hearing ability from 80Hz to 500Hz. In addition, the auditory thresholds over 800Hz increased rapidly. The mean auditory thresholds of the Redlip croaker at the test frequencies from 80Hz to 800Hz were 90.7dB, 93.4dB, 92.9dB, 94.4dB, 95.5dB and 108dB, respectively. Auditory masking for the redlip croaker was measured using masking stimuli with the spectrum level range of about 66, 71, 75dB (0dB re $1{\mu}Pa/{\sqrt{Hz}}$). According to white noise level, the auditory thresholds increased as compared with thresholds in a quiet background noise. The Auditory masking by the white noise spectrum level was stared over about 70dB within 80~500Hz. Critical ratio ranged from minimum 20.7dB to maximum 25.5dB at test frequencies of 80Hz~500Hz.

• Journal of Korea Multimedia Society
• /
• v.5 no.4
• /
• pp.405-414
• /
• 2002

In this paper, we propose a new objective speech quality measure using noise masking threshold for speech quality assessment of mobile telephone network environments, and verify the effectiveness of the proposed method through the experiments. For such a purpose, well known objective speech quality measures such as BSD and PSQM are first evaluated for digital mobile telephone network environments. However, these conventional methods does not have good performance under mobile networks environments compared to literary results. To be mote effective objective speech quality measure under mobile telephone environments, the proposed method employs human psychoacoustic masking effect. The DMOS, instead of MOS, is used as a subjective speech quality measure for performance evaluation. The performance comparison are carried out with speech data collected from digital mobile telephone environments. As results, the proposed measure have and average 4% higher performance, in terms of correlation, than existing objective speech quality measures such as BSD and PSQM.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.21 no.2
• /
• pp.149-156
• /
• 2011

In the recent years, power analysis attacks were widely investigated, and so various countermeasures have been proposed. In the case of block ciphers, masking methods that blind the intermediate values in the en/decryption computations are well-known among these countermeasures. But the cost of non-linear part is extremely high in the masking method of block cipher, and so the countermeasure for S-box must be efficiently constructed in the case of AES, ARIA and SEED. Existing countermeasures for S-box use the masked S-box table to require 256 bytes RAM corresponding to one S-box. But, the usage of the these countermeasures is not adequate in the lightweight security devices having the small size of RAM. In this paper, we propose the new countermeasure not using the masked S-box table to make up for this weak point. Also, the new countermeasure reduces time-complexity as well as the usage of RAM because this does not consume the time for generating masked S-box table.

• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.167-175
• /
• 2021

Semi-supervised learning uses both labeled data and unlabeled data. Recently consistency regularization is very popular in semi-supervised learning. Unsupervised data augmentation (UDA) that uses unlabeled data augmentation is also based on the consistency regularization. The Kullback-Leibler divergence is used for the loss of unlabeled data and cross-entropy for the loss of labeled data through UDA learning. UDA uses techniques such as training signal annealing (TSA) and confidence-based masking to promote performance. In this study, we propose to use Jensen-Shannon divergence instead of Kullback-Leibler divergence, reverse-TSA and not to use confidence-based masking for performance improvement. Through experiment, we show that the proposed technique yields better performance than those of UDA.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.5
• /
• pp.807-816
• /
• 2022

See-In-The-Middle (SITM) is an analysis technique that uses Side-Channel information for differential cryptanalysis. This attack collects unmasked middle-round power traces when implementing block ciphers to select plaintext pairs that satisfy the attacker's differential pattern and utilize them for differential cryptanalysis to recover the key. Romulus, one of the final candidates for the NIST Lightweight Cryptography standardization competition, is based on Tweakable block cipher Skinny-128-384+. In this paper, the SITM attack is applied to Skinny-128-384 implemented with 14-round partial masking. This attack not only increased depth by one round, but also significantly reduced the time/data complexity to 2^14.93/2^14.93. Depth refers to the round position of the block cipher that collects the power trace, and it is possible to measure the appropriate number of masking rounds required when applying the masking technique to counter this attack. Furthermore, we extend the attack to Romulus's Nonce-based AE mode Romulus-N, and Tweakey's structural features show that it can attack with less complexity than Skinny-128-384.