Myeongjin Kwak, Jungwon Lee, Hyoju Seo, Mingyu Sung, Yongtae Kim
{"title":"Training and Inference using Approximate Floating-Point Arithmetic for Energy Efficient Spiking Neural Network Processors","authors":"Myeongjin Kwak, Jungwon Lee, Hyoju Seo, Mingyu Sung, Yongtae Kim","doi":"10.1109/ICEIC51217.2021.9369724","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369724","url":null,"abstract":"This paper presents a systematic analysis of spiking neural network (SNN) performance with reduced computation precisions using approximate adders. We propose an IEEE 754-based approximate floating-point adder that applies to the leaky integrate-and-fire (LIF) neuron-based SNN operation for both training and inference. The experimental results under a two-layer SNN for MNIST handwritten digit recognition application show that 4-bit exact mantissa adder with 19-bit approximation for lower-part OR adder (LOA), instead of 23-bit full-precision mantissa adder, can be exploited to maintain good classification accuracy. When adopted LOA as mantissa adder, it can achieve up to 74.1% and 96.5% of power and energy saving, respectively.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114942336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Development and RSSI Testing of IoT Ambient Device for Building Energy Management System","authors":"Tanakorn Inthasuth, Pongsatit Chuayniam","doi":"10.1109/ICEIC51217.2021.9369746","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369746","url":null,"abstract":"This paper present the development and testing of ambient measuring device with the Wi-Fi connection with the Network Platform Internet of Every Thing (NETPIE). The device is able to apply for the building energy management system. The system is consist of two parts, which are sensors and NETPIE platform; both are connected via internet in MQTT protocol. This become the guideline for Internet of Things (IoT) device installation in building that is connected via internet by Wi-Fi. For the testing, we measures the receive signal strange indicator (RSSI) at three separated areas, including staff room, terrace and elevator front. The result shows the lowest signal is at elevator front because of distance from the switch hub and obstacles. However, this experiment show that all data are transfers. Furthermore, the result can be used for designing of Wi-Fi network installation for IoT devices.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131354891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Design and Implementation of Ternary Carry Lookahead Adder on FPGA","authors":"Jaeyoon Park, Youngmin Kim","doi":"10.1109/ICEIC51217.2021.9369710","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369710","url":null,"abstract":"Ternary value logic (TVL) has advantages over binary system, such as providing smaller chip area and faster computation speed. However, ternary hardware implementation is in theoretical, simulation levels. Moreover, no hardware description language for ternary logic is developed. In this paper, by representing 1 trit (trinary digit) with 2 bit, ternary logic is implemented and analyzed in FPGA. To specify the performance of TVL, ternary carry lookahead adder is implemented on FPGA and the speed and power dissipation of the arithmetic unit are measured. The performance of 21-trit ternary CLA is compared with 32-bit binary CLA. The results show that ternary CLA is faster up to 10.36% and consumes 13.54% less power than binary CLA. In addition, performances of ripple carry adder using both ternary and binary logic are simulated.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126731060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Integrating GPS Sensor and Beamforming for UAV Tracking Algorithm","authors":"Ha-Lim Song, Young-Chai Ko","doi":"10.1109/ICEIC51217.2021.9369823","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369823","url":null,"abstract":"Integrating flight sensor data into beamforming technology in unmanned aerial vehicles (UAVs) communications is an attractive solution. Due to the movements and perturbations of UAV, maintaining beam alignment is a challenging issue. In this paper, we propose an integrated flight sensor data and beamforming algorithm for UAV tracking. We can reduce the beam training overhead drastically with the global positioning system (GPS). However, due to relatively large errors and a long period of the GPS signal data, we consider exploiting beamforming to persist the beam alignment even when the GPS signal is absent. The simulation results show that the proposed scheme can improve the performance compared to a GPS-single system. In addition, the proposed algorithm can accomplish high beamforming gain with less beam training overhead compared to the conventional beamforming algorithm.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123012873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Quantitative Analysis and Simple Monitoring for Partial Discharge from Transformer","authors":"S. Kwak, Y. Kim","doi":"10.1109/ICEIC51217.2021.9369769","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369769","url":null,"abstract":"The various signals are output such as partial discharge, vibration, ultrasonic waves, and acoustic emission when the transformers have deformation or abnormality. These signals need to be monitored since they are important signals to notice the transformer failure. To prevent this, the monitoring system of transformer uses a high-speed data acquisition system to extract various physical quantities from the raw abnormal signal. Among the types of abnormal signals, partial discharge (PD) is caused by failure of the insulator, which has a great influence on the life and maintenance of the transformer. In this paper, a PD signal is mimicked by using a PD generator to analyze the partial discharge of a transformer. PD signal is simply analyzed by fast Fourier transform. Also, by proposing a simple monitoring system, it is introduced that a communication method of root-mean-square (RMS) value, maximum value, and maximum RMS value.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121243708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Current Differencing Buffered Amplifier Based Memristive Quadrature Oscillator","authors":"Ankur Singh, S. Borah, M. Ghosh","doi":"10.1109/ICEIC51217.2021.9369780","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369780","url":null,"abstract":"This article investigates the practicability of using the attributes of a Current Differencing Buffered Amplifier (CDBA) as an active element and Memristors to achieve sustained quadrature oscillations with low-power supply. The periodical oscillations are obtained by employing three memristors and two grounded capacitors. The performance characteristics of Memristor based quadrature oscillator preferably conventional resistor contributes more desirable oscillator characteristics are examined here. PSPICE simulation outcomes are introduced to validate the theoretical propositions that provide independent control of frequency and condition of oscillation. The influence of parasitic reduced due to the internally grounded ends of CDBA and grounded capacitors. The other performance analysis such as non-ideality, sensitivity, percentage of total harmonic distortion (%THD) and Monte-Carlo analysis are also discussed for the proposed structure. The total power consumption of the proposed configuration is 0.917mW. In addition, an application of the proposed MQO configuration is presented.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116178797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Analysis of the Smartphone Camera Exposure Effect on Laser Extraction","authors":"L. Pham, Duong Nguyen-Ngoc Tran, C. Rhie, J. Jeon","doi":"10.1109/ICEIC51217.2021.9369723","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369723","url":null,"abstract":"In this study, we investigate the effect of smartphone camera exposure on laser extraction performance in a digital image. The laser extraction is the most crucial operation in laser triangulation systems that obtain object measurements from the projected laser light captured in the image. The exposure settings consist of sensor sensitivity (ISO), aperture, and shutter speed. This study provides insight into the smartphone settings and the exposure adjustment for further complicated laser extraction. A smartphone camera (Apple iPhone 11) and a laser stripe projector operating at 405 nm were employed. Analysis of different ISO values, the effect of aperture and shutter speed on the laser extraction quality was performed.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124733159","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sangheum Lee, Taehyeon Kim, Myunghyun Kim, Donghan Kim
{"title":"Simultaneous control algorithm for mobile manipulator using MMPPE","authors":"Sangheum Lee, Taehyeon Kim, Myunghyun Kim, Donghan Kim","doi":"10.1109/ICEIC51217.2021.9369735","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369735","url":null,"abstract":"This paper proposes the simultaneous control algorithm using Mobile Manipulator Pre-Pose Estimator (MMPPE). A conventional method is a sequential method in which the manipulator is controlled after the mobile platform moves. However, the proposed method recognizes the target pose in advance and predicts the posture of the manipulator and the mobile robot. It has more time efficiency than conventional sequential control method. By integrating this based on ROS and verify the algorithm using simulation.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"158 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122979098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hatem Ibrahem, Ahmed Salem, Bilel Yagoub, Hyun-Soo Kang
{"title":"CCC: Weakly Supervised Semantic Segmentation using Class Contribution Classification","authors":"Hatem Ibrahem, Ahmed Salem, Bilel Yagoub, Hyun-Soo Kang","doi":"10.1109/ICEIC51217.2021.9369721","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369721","url":null,"abstract":"We propose a new semantic Segmentation technique that depends on the convolutional neural network of Xception using a modified categorical cross-entropy function to perform class contribution classification. We introduce a class contribution function for each class contribution in the image instead of categorical cross-entropy for maximum class classification. This approach is to be used for weakly supervised semantic segmentation (WSSS) using image-level annotation instead of the pixel-level annotations used in fully supervised semantic segmentation. We trained and tested our approach on both PASCAL VOC2007 and VOC2012 datasets. We show that our approach outperformed many other weakly supervised methods, specifically, it can attain a segmentation accuracy of 49.7% on PASCAL VOC2007 test set and an accuracy of 41.3% on PASCAL VOC2012 test set while the network is trained to learn the semantics of the objects in the image using the cheap image label annotation instead of the expensive and time-consuming segmentation masks.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122787670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Improvement of worker scaling-based scheduling algorithm to efficiently respond to explosion of micro function service requests","authors":"Youngho Kim, Gyuil Cha","doi":"10.1109/ICEIC51217.2021.9369745","DOIUrl":"https://doi.org/10.1109/ICEIC51217.2021.9369745","url":null,"abstract":"We present improved worker scheduler algorithm and architecture to minimize the scheduling cost and function service response time of FaaS platform. We try to solve the problem of performance degradation of previous scaling-based worker scheduling algorithm due to the explosion of the function invocation requests and the blocking-based communication channel thread with the computation node. To this end, two algorithms including Load-based function worker demand prediction algorithm and adoption of asynchronous communication channel integration and non-blocking allocation method have been developed. Through the performance evaluation of proposed worker scheduling algorithm on FaaS worker scheduler, the proposed worker scheduler achieves speedups as high as 10.1 and 16.8 compared to the baseline Bare-metal Computing nodes with the CPU and GPU devices using the in-house TSL Inferencing application.","PeriodicalId":170294,"journal":{"name":"2021 International Conference on Electronics, Information, and Communication (ICEIC)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125427659","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}