• International Journal of Computer Science & Network Security
• /
• v.22 no.9
• /
• pp.175-182
• /
• 2022

In our daily life, we come across different types of information, for example in the format of multimedia and text. We all need different types of information for our common routines as watching/reading the news, listening to the radio, and watching different types of videos. However, sometimes we could run into problems when a certain type of information is required. For example, someone is listening to the radio and wants to listen to jazz, and unfortunately, all the radio channels play pop music mixed with advertisements. The listener gets stuck with pop music and gives up searching for jazz. So, the above example can be solved with an automatic audio classification system. Deep Learning (DL) models could make human life easy by using audio classifications, but it is expensive and difficult to deploy such models at edge devices like nano BLE sense raspberry pi, because these models require huge computational power like graphics processing unit (G.P.U), to solve the problem, we proposed DL model. In our proposed work, we had gone for a low complexity model for Audio Event Detection (AED), we extracted Mel-spectrograms of dimension 128×431×1 from audio signals and applied normalization. A total of 3 data augmentation methods were applied as follows: frequency masking, time masking, and mixup. In addition, we designed Convolutional Neural Network (CNN) with spatial dropout, batch normalization, and separable 2D inspired by VGGnet [1]. In addition, we reduced the model size by using model quantization of float16 to the trained model. Experiments were conducted on the updated dataset provided by the Detection and Classification of Acoustic Events and Scenes (DCASE) 2020 challenge. We confirm that our model achieved a val_loss of 0.33 and an accuracy of 90.34% within the 132.50KB model size.

• Smart Media Journal
• /
• v.12 no.4
• /
• pp.19-26
• /
• 2023

IoT edge services utilizing neuromorphic hardware architectures are suitable for autonomous IoT applications as they perform intelligent processing on the device itself. However, spiking neural networks applied to neuromorphic hardware are difficult for IoT developers to comprehend due to their complex structures and various hyper-parameters. In this paper, we propose a method for generating spiking neural network (SNN) models that satisfy user performance requirements while considering the constraints of neuromorphic hardware. Our proposed method utilizes previously trained models from pre-processed data to find optimal SNN model parameters from profiling data. Comparing our method to a naive search method, both methods satisfy user requirements, but our proposed method shows better performance in terms of runtime. Additionally, even if the constraints of new hardware are not clearly known, the proposed method can provide high scalability by utilizing the profiled data of the hardware.

• Smart Media Journal
• /
• v.11 no.2
• /
• pp.70-76
• /
• 2022

IoT application development using a cloud server causes problems such as data transmission and reception delay, network traffic, and cost for real-time processing support in network connected hardware. To solve this problem, edge cloud-based platforms can use neuromorphic hardware to enable fast data transfer. In this paper, we propose a model optimization method for supporting spiking neural networks on FPGA hardware. We focused on auto-adjusting network model parameters optimized for neuromorphic hardware. The proposed method performs optimization to show higher performance based on user requirements for accuracy. As a result of performance analysis, it satisfies all requirements of accuracy and showed higher performance in terms of expected execution time, unlike the naive method supported by the existing open source framework.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.6
• /
• pp.61-67
• /
• 2023

The rise of 5G and the proliferation of smart devices have underscored the significance of multi-access edge computing (MEC). Amidst this trend, interest in effectively processing computation-intensive and latency-sensitive applications has increased. This study investigated a novel task offloading strategy considering the probabilistic MEC environment to address these challenges. Initially, we considered the frequency of dynamic task requests and the unstable conditions of wireless channels to propose a method for minimizing vehicle power consumption and latency. Subsequently, our research delved into a deep reinforcement learning (DRL) based offloading technique, offering a way to achieve equilibrium between local computation and offloading transmission power. We analyzed the power consumption and queuing latency of vehicles using the deep deterministic policy gradient (DDPG) and deep Q-network (DQN) techniques. Finally, we derived and validated the optimal performance enhancement strategy in a vehicle based MEC environment.

• Journal of Internet of Things and Convergence
• /
• v.10 no.2
• /
• pp.17-23
• /
• 2024

Losses in domestic water supply due to leaks are very large, such as fractures and defects in pipelines. Therefore, preventive measures to prevent water leakage are necessary. We propose the development of a leakage detection sensor utilizing vibration sensors and present an optimal leakage detection algorithm leveraging artificial intelligence. Vibrational sound data acquired from water pipelines undergo a preprocessing stage using FFT (Fast Fourier Transform), followed by leakage classification using an optimized tree-based boosting algorithm. Applying this method to approximately 260,000 experimental data points from various real-world scenarios resulted in a 97% accuracy, a 4% improvement over existing SVM(Support Vector Machine) methods. The processing speed also increased approximately 80 times, confirming its suitability for edge device applications.

• Smart Structures and Systems
• /
• v.33 no.6
• /
• pp.399-414
• /
• 2024

This study discusses the long-term deformation monitoring and shape sensing of bridge girder surfaces with an automated extraction scheme for point clouds in the Region Of Interest (ROI), invariant to the position of a Light Detection And Ranging system (LiDAR). Advanced smart construction necessitates continuous monitoring of the deformation and shape of bridge girders during the construction phase. An automated scheme is proposed for reconstructing geometric model of ROI in the presence of noisy non-stationary background. The proposed scheme involves (i) denoising irrelevant background point clouds using dimensions from the design model, (ii) extracting the outer boundaries of the bridge girder by transforming and processing the point cloud data in a two-dimensional image space, (iii) extracting topology of pre-defined targets using the modified Otsu method, (iv) registering the point clouds to a common reference frame or design coordinate using extracted predefined targets placed outside ROI, and (v) defining the bounding box in the point clouds using corresponding dimensional information of the bridge girder and abutments from the design model. The surface-fitted reconstructed geometric model in the ROI is superposed consistently over a long period to monitor bridge shape and derive deflection during the construction phase, which is highly correlated. The proposed scheme of combining 2D-3D with the design model overcomes the sensitivity of 3D point cloud registration to initial match, which often leads to a local extremum.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.3
• /
• pp.1645-1651
• /
• 2015

This paper describes the changes of shape accuracy in workpiece materials depending on the turning clearance angle. The experiments started from choosing three workpiece materials, SM45C(machine structural carbon steel), STS303(stainless steel) and SCM415 (chrome-molybdenum steel). The experiments showed specifically how features of selected materials changed when they were processed with diverse machining depths, 0.1 mm, 0.2 mm and 0.3 mm, with various negative angles, $0.0^{\circ}(-6.0^{\circ})$, $0.3^{\circ}(-6.3^{\circ})$ and $0.9^{\circ}(-6.9^{\circ})$, and called cutting edge inclination starting from a fixed rotational speed, 2,500 rpm, focusing on the feed rate, 0.07 mm/rev and 0.10 mm/rev. The results of the accuracy of processing, cylindricity, deviation from coaxiality, etc. were compared using the graph and table. The accuracy of cylindricity in the order of degree $0.0^{\circ}{\rightarrow}0.3^{\circ}{\rightarrow}0.9^{\circ}$ depending on the workpiece materials showed the best cylindricity when it was $0.9^{\circ}$. In conclusion, the accuracy improved in specific degrees irrespective of the quality of the materials when the bite negative angles increased. This means that workability improved in these experiments. In addition, the processing shape changed depending on depth of the cut and feed rate.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.12
• /
• pp.17-26
• /
• 1998

Ultrasonic sensors are widely used in various applications due to advantages of low cost, simplicity in construction, mechanical robustness, and little environmental restriction in usage. But for the application of object recognition, ultrasonic sensors exhibit several shortcomings of poor directionality which results in low spatial resolution of objects, and specularity which gives frequent erroneous range readings. The time-of-flight(TOF) method generally used for distance measurement can not distinguish small object patterns of plane, corner or edge. To resolve the problem, an increased number of the sensors in the forms of a linear array or 2-dimensional array of the sensors has been used. Also better resolution has been obtained by shifting the array in several steps using mechanical actuators. Also simple patterns are classified based on analyzing signal reflections. In this paper we propose a method of a sensor array system with improved capability in pattern distinction using electronic circuits accompanying the sensor array, and intelligent algorithm based on neuro-fuzzy processing of data fusion. The circuit changes transmitter output voltages of array elements in several steps. A set of different return signals from neighborhood sensors is manipulated to provide enhanced pattern recognition in the aspects of inclination angle, size and shift as well as distance of objects. The results show improved resolution of the measurements for smaller targets.

• 전자공학회논문지 IE
• /
• v.46 no.4
• /
• pp.57-65
• /
• 2009

This paper presents the image contrast enhancement technique suitable for local dimming backlight of small-sized mobile display while achieving the reduction of the power consumption. In addition to the large-sized TFT-LCD, small-sized one has adopted LED for backlight. Since, conventionally, LED was mounted on the side edge of a display panel, global dimming method has been widely used. However, recently, new advanced method of local dimming by placing the LED to the backside of the display panel and it raised the necessity of sub-blocked processing after partitioning the target image. When the sub-blocked image has low brightness, the supply current of a backlight LED is reduced, which gives both enhancement of contrast ratio and power consumption reduction. In this paper, we propose simple and improved image enhancement algorithm suitable for the small-sized mobile display. After partitioning the input image by equal sized blocks and analyzing the pixel information in each block, we realize the primary contrast enhancement by independently processing the sub-blocks using the information such as histogram, mean, and standard deviation values of luminance(Y) component. And then resulting information is transferred to each backlight control unit for local dimming to realize the secondary contrast enhancement as well as reduction of power consumption.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.45 no.3
• /
• pp.12-19
• /
• 2008

Detection of a number plate consists of three stages; division of a number plate, extraction of each character from the plate, recognition of the characters. Among of these three states, division stage of a number plate is the most important part and also the most time-consuming state. This paper suggests an effective region extraction method of a number plate for various images obtained from unmanned inspection systems of illegal parking violation, especially when we have to consider the diverse surrounding environments of roads. Our approaching method detects each region by investigating the characteristics in changes of brightness and intensity between the background part and character part, and the characteristics on character parts such as the sizes, heights, widths, and distance in between two characters. The method also divides a number plate into different types of the plate. This research can solve the number plate region detection failure problems caused by plate edge damages not only for Korean domestic number plates but also for new European style number plates. The method also reduces the time consumption by processing the detection in real-time, therefore, it can be used as a practical solution.