• Journal of Computing Science and Engineering
• /
• v.12 no.1
• /
• pp.24-35
• /
• 2018

Deep convolutional neural network (DCNN) is an advanced technology in image recognition. Because of extreme computing resource requirements, DCNN implementation with software alone cannot achieve real-time requirement. Therefore, the need to implement DCNN accelerator hardware is increasing. In this paper, we present a field programmable gate array (FPGA)-based hardware accelerator design of DCNN targeting handwritten Hangul character recognition application. Also, we present design optimization techniques in SDAccel environments for searching the optimal FPGA design space. The techniques we used include memory access optimization and computing unit parallelism, and data conversion. We achieved about 11.19 ms recognition time per character with Xilinx FPGA accelerator. Our design optimization was performed with Xilinx HLS and SDAccel environment targeting Kintex XCKU115 FPGA from Xilinx. Our design outperforms CPU in terms of energy efficiency (the number of samples per unit energy) by 5.88 times, and GPGPU in terms of energy efficiency by 5 times. We expect the research results will be an alternative to GPGPU solution for real-time applications, especially in data centers or server farms where energy consumption is a critical problem.

• Smart Media Journal
• /
• v.11 no.3
• /
• pp.26-30
• /
• 2022

Spiking neural network is a neural network that applies the working principle of real brain neurons. Due to the biological mechanism of neurons, it consumes less power for training and reasoning than conventional neural networks. Recently, as deep learning models become huge and operating costs increase exponentially, the spiking neural network is attracting attention as a third-generation neural network that connects convolution neural networks and recurrent neural networks, and related research is being actively conducted. However, in order to apply the spiking neural network model to the industry, a lot of research still needs to be done, and the problem of model retraining to apply a new model must also be solved. In this paper, we propose a method to minimize the cost of model retraining by extracting the weights of the existing trained deep learning model and converting them into the weights of the spiking neural network model. In addition, it was found that weight conversion worked correctly by comparing the results of inference using the converted weights with the results of the existing model.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.4
• /
• pp.685-692
• /
• 2024

In this paper, we propose a new CNN model to improve the recognition performance of command voice signals. This method obtains a spectrogram image after performing a short-time Fourier transform (STFT) of the input signal and improves command recognition performance through supervised learning using a CNN model. After Fourier transforming the input signal for each short-time section, a spectrogram image is obtained and multi-classification learning is performed using a CNN deep learning model. This effectively classifies commands by converting the time domain voice signal to the frequency domain to express the characteristics well and performing deep learning training using the spectrogram image for the conversion parameters. To verify the performance of the speech recognition system proposed in this study, a simulation program using Tensorflow and Keras libraries was created and a simulation experiment was performed. As a result of the experiment, it was confirmed that an accuracy of 92.5% could be obtained using the proposed deep learning algorithm.

• Korean Journal of Radiology
• /
• v.24 no.8
• /
• pp.807-820
• /
• 2023

Objective: To assess whether computed tomography (CT) conversion across different scan parameters and manufacturers using a routable generative adversarial network (RouteGAN) can improve the accuracy and variability in quantifying interstitial lung disease (ILD) using a deep learning-based automated software. Materials and Methods: This study included patients with ILD who underwent thin-section CT. Unmatched CT images obtained using scanners from four manufacturers (vendors A-D), standard- or low-radiation doses, and sharp or medium kernels were classified into groups 1-7 according to acquisition conditions. CT images in groups 2-7 were converted into the target CT style (Group 1: vendor A, standard dose, and sharp kernel) using a RouteGAN. ILD was quantified on original and converted CT images using a deep learning-based software (Aview, Coreline Soft). The accuracy of quantification was analyzed using the dice similarity coefficient (DSC) and pixel-wise overlap accuracy metrics against manual quantification by a radiologist. Five radiologists evaluated quantification accuracy using a 10-point visual scoring system. Results: Three hundred and fifty CT slices from 150 patients (mean age: 67.6 ± 10.7 years; 56 females) were included. The overlap accuracies for quantifying total abnormalities in groups 2-7 improved after CT conversion (original vs. converted: 0.63 vs. 0.68 for DSC, 0.66 vs. 0.70 for pixel-wise recall, and 0.68 vs. 0.73 for pixel-wise precision; P < 0.002 for all). The DSCs of fibrosis score, honeycombing, and reticulation significantly increased after CT conversion (0.32 vs. 0.64, 0.19 vs. 0.47, and 0.23 vs. 0.54, P < 0.002 for all), whereas those of ground-glass opacity, consolidation, and emphysema did not change significantly or decreased slightly. The radiologists' scores were significantly higher (P < 0.001) and less variable on converted CT. Conclusion: CT conversion using a RouteGAN can improve the accuracy and variability of CT images obtained using different scan parameters and manufacturers in deep learning-based quantification of ILD.

• Journal of Power System Engineering
• /
• v.20 no.1
• /
• pp.11-17
• /
• 2016

In this study, the simulation for performance comparison between basic single stage organic rankine cycle, multi stage reheater cycle and multi stage reheater & recuperator cycle was carried out. The multi stage reheater cycle and multi stage reheater & recuperator cycle was designed to improve the efficiency for organic rankine cycle using heat source from industrial waste heat and heat sink from deep ocean water. R245fa was selected as a refrigerant for the cycle and system efficiencies were simulated by the variation of the heat sink temperature and the cycle classification. Performance characteristics were simulated by using the Aspen HYSYS. It was confirmed that the system efficiency was decreased by the increase of heat sink temperature. These results can be considered to be applied as geo-ocean thermal energy conversion in where plenty of geothermal or ocean thermal resource exist.

• Journal of Microbiology and Biotechnology
• /
• v.13 no.4
• /
• pp.578-583
• /
• 2003

The desulfurization genes (dszABC) were cloned from Gordonia nitida. Nucleotide sequences similarity between the dszABC genes of G. nitida and those of Rhodococcus rhodochrous IGTS8 was 89%. The similarities of deduced amino acids between the two were 86% for DszA, 86% for DszB, and 90% for DszC. The G. nitida dszABC genes were expressed in several different Escherichia coli strains under an inducible trc promoter. Cultivation of these metabolically engineered E. coli strains in the presence of 0.2 mM dibenzothiophene (DBT) allowed the conversion of DBT to 2-hydroxybiphenyl (2-HBP), which is the final metabolite of the sulfur-specific desulfurization pathway. The maximum conversion of DBT to 2-HBP was 16% in 60 h. Recombinant E. coli was applied for the deep desulfurization of diesel oil supplemented into the medium at 5% (v/v). Sulfur content in diesel oil was decreased from 250 mg sulfur/1 to 212.5 mg sulfur/1, resulting in the removal of 15% of sulfur in diesel oil in 60 h.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.6
• /
• pp.3029-3045
• /
• 2017

The main processes of a cognitive radio system include spectrum sensing, spectrum decision, spectrum sharing, and spectrum conversion. Experimental results show that these stages introduce a time delay that affects the spectrum sensing accuracy, reducing its efficiency. To reduce the time delay, the frequency spectrum prediction was proposed to alleviate the burden on the spectrum sensing. In this paper, the deep recurrent neural network (DRNN) was proposed to predict the spectrum of multiple time slots, since the existing methods only predict the spectrum of one time slot. The continuous state of a channel is divided into a many time slots, forming a time series of the channel state. Since there are more hidden layers in the DRNN than in the RNN, the DRNN has fading memory in its bottom layer as well as in the past input. In addition, the extended Kalman filter was used to train the DRNN, which overcomes the problem of slow convergence and the vanishing gradient of the gradient descent method. The spectrum prediction based on the DRNN was verified with a WiFi signal, and the error of the prediction was analyzed. The simulation results proved that the multiple slot spectrum prediction improved the spectrum efficiency and reduced the energy consumption of spectrum sensing.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.3
• /
• pp.35-41
• /
• 2013

An efficient multi-stage rectifier for wireless power transfer in deep implant medical devices is implemented using $0.18-{\mu}m$ CMOS technology. The presented three-stage rectifier employs a cross-coupled topology to boost a small input AC signal from the external device to produce a 1.2-1.5 V output DC signal for the implant device. The designed rectifier achieves a maximum measured power conversion efficiency of 70% at 13.56 MHz under the conditions of a low 0.6-Vpp RF input signal with a $10-k{\Omega}$ output load resistance.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2001.11b
• /
• pp.497-500
• /
• 2001

Due to their better photosensitivity in X-ray, the amorphous selenium based photoreceptor is widely used on the X-ray conversion materials. It was possible to control the charge carrier transport of amorphous selenium by suitably alloying a-Se with other elements(e.g. As, Cl). The charge transport properties of amorphous Selenium is decided on hole which is induced from metal to selenium in metal-selenium junction and which is transferred in a-Se bulk. This phenomenon is resulted of changing electric field owing to increasing of space charge by deep trap of a-Se bulk. In this paper, We dopped the chlorine to compensate deep hole trap and deposited blocking layer using dielectric material to prevent from increasing space charge for injection charge between metal electrode and a-Se layer. We compared space charge and the decreasing of trap density through measuring dark and photo current. 缀Ѐ㘰〻ሀ䝥湥牡氠瑥捨湯汯杹

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2002.11a
• /
• pp.656-661
• /
• 2002

Multipass Ti:sapphire amplifier for the light source of lidar was developed in an angular-multiplexing, and the characteristics of output energy and spectrum was investigated. In the two-stage multipass amplifier, we obtained the maximum output energy of 42 mJ, the amplification gain of 21 dB and the output efficiency of 26% on the wavelength of 790 nm. In the tuning range of 715~930nm the spectral linewidth is 0.05 $cm^{-1}$ /. The conversion efficiencies of 35% for SHG at 780 m and 13% for THG at 390 nm are obtained respectively. The continuous tunabilities of 240~306 m UV region and 360~460 nm in deep-blue region could be achieved.