Seong-Won Lee
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Seong-Won Lee has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Ho-Young Kim
Department of Electrical Engineering, University of Colorado Denver, Denver, CO 80204 USA
3 shared publications - Jo-In Kang
Department of AI Convergence Engineering, Kwangwoon University, Seoul, South Korea
3 shared publications Seung-Hyun Choi
Department of Computer Engineering, Kwangwoon University
3 shared publicationsYong Ho Song
Hanyang University
2 shared publications- Chan-Ho Moon
Hanwha VisionNext, Seongnam, South Korea
1 shared publication - Ching-Jan Chen
Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan
1 shared publication
Latest Publications
- Open Access
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 76773-76788
SCOPE: Stochastic Collaborative Learning via Peer Exchange for Fully Decentralized Non-IID IoT Networks
Decentralized learning across Internet of Things (IoT) and Wireless Sensor Network (WSN) nodes often suffers from unstable convergence under non-IID data and intermittent connectivity. This paper presents Stochastic Collaborative Learning via Peer Exchange (SCOPE), a fully decentralized learning designed to achieve stable and fair convergence without any central coordinator. SCOPE integrates divergence-aware peer pairing based on Jensen–Shannon Divergence (JSD), stability-guided batch size adaptation, and proxy-data–based model validation into a unified decentralized learning cycle. Through this integrated design, SCOPE promotes directional knowledge exchange, balances exploration and convergence during local training, and suppresses the propagation of biased updates. Experiments on clustered non-IID datasets, conducted on both a Jetson Nano testbed and an MPI-based simulator, demonstrate that SCOPE consistently improves convergence smoothness, inter-node fairness, and final accuracy compared to existing decentralized baselines. These results indicate that SCOPE provides a statistically grounded and scalable solution for stable decentralized learning in resource-constrained IoT environments.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 76773-76788
Open AccessSCOPE: Stochastic Collaborative Learning via Peer Exchange for Fully Decentralized Non-IID IoT Networks
Decentralized learning across Internet of Things (IoT) and Wireless Sensor Network (WSN) nodes often suffers from unstable convergence under non-IID data and intermittent connectivity. This paper presents Stochastic Collaborative Learning via Peer Exchange (SCOPE), a fully decentralized learning designed to achieve stable and fair convergence without any central coordinator. SCOPE integrates divergence-aware peer pairing based on Jensen–Shannon Divergence (JSD), stability-guided batch size adaptation, and proxy-data–based model validation into a unified decentralized learning cycle. Through this integrated design, SCOPE promotes directional knowledge exchange, balances exploration and convergence during local training, and suppresses the propagation of biased updates. Experiments on clustered non-IID datasets, conducted on both a Jetson Nano testbed and an MPI-based simulator, demonstrate that SCOPE consistently improves convergence smoothness, inter-node fairness, and final accuracy compared to existing decentralized baselines. These results indicate that SCOPE provides a statistically grounded and scalable solution for stable decentralized learning in resource-constrained IoT environments.
- Open Access
- 25 October 2024
- Published by MDPI AG in Electronics
- Vol. 13 (21) , 4193
RNEP: Random Node Entropy Pairing for Efficient Decentralized Training with Non-IID Local Data
The proliferation of edge devices and advancements in Internet of Things technology have created a vast array of distributed data sources, necessitating machine learning models that can operate closer to the point of data generation. Traditional centralized machine learning approaches often struggle with real-time big data applications, such as climate prediction and traffic simulation, owing to high communication costs. In this study, we introduce random node entropy pairing, a novel distributed learning method for artificial neural networks tailored to distributed computing environments. This method reduces communication overhead and addresses data imbalance by enabling nodes to exchange only the weights of their local models with randomly selected peers during each communication round, rather than sharing the entire dataset. Our findings indicate that this approach significantly reduces communication costs while maintaining accuracy, even when learning from non-IID local data. Furthermore, we explored additional learning strategies that enhance system accuracy by leveraging the characteristics of this method. The results demonstrate that random node entropy pairing is an effective and efficient solution for distributed learning in environments where communication costs and data distribution present significant challenges.
- 25 October 2024
- Published by MDPI AG in Electronics
- Vol. 13 (21) , 4193
Open AccessRNEP: Random Node Entropy Pairing for Efficient Decentralized Training with Non-IID Local Data
The proliferation of edge devices and advancements in Internet of Things technology have created a vast array of distributed data sources, necessitating machine learning models that can operate closer to the point of data generation. Traditional centralized machine learning approaches often struggle with real-time big data applications, such as climate prediction and traffic simulation, owing to high communication costs. In this study, we introduce random node entropy pairing, a novel distributed learning method for artificial neural networks tailored to distributed computing environments. This method reduces communication overhead and addresses data imbalance by enabling nodes to exchange only the weights of their local models with randomly selected peers during each communication round, rather than sharing the entire dataset. Our findings indicate that this approach significantly reduces communication costs while maintaining accuracy, even when learning from non-IID local data. Furthermore, we explored additional learning strategies that enhance system accuracy by leveraging the characteristics of this method. The results demonstrate that random node entropy pairing is an effective and efficient solution for distributed learning in environments where communication costs and data distribution present significant challenges.
- Open Access
- 8 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 156663-156671
A Random Modulation Spread-Spectrum Digital PWM for a Low System Clock Digital Buck Converter
The need for a high-efficiency power system has emerged since many IoT devices are mainly operated by batteries. A power system with analog control can have a high control bandwidth, but it is relatively vulnerable to noise disturbances and may suffer from control problems. On the other hand, a digitally controlled power system has high reliability and great controllability at the expense of switching noise and EMI issues. To implement a digitally controlled PWM with high resolution and low EMI, this study proposes a random carrier frequency modulation with dithered duty cycle method, where wide spectrum dithering and random modulation of spread spectrum PWM are tightly integrated. The proposed system was not only able to suppress voltage ripples, but also possible to minimize the occurrence of dependent frequencies due to the fixed dithering pattern. The spectrum of the output voltage showed that the switching frequency component was reduced by about 6.1dB and the output voltage ripples was also reduced compared to the existing PWM techniques.
- 8 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 156663-156671
Open AccessA Random Modulation Spread-Spectrum Digital PWM for a Low System Clock Digital Buck Converter
The need for a high-efficiency power system has emerged since many IoT devices are mainly operated by batteries. A power system with analog control can have a high control bandwidth, but it is relatively vulnerable to noise disturbances and may suffer from control problems. On the other hand, a digitally controlled power system has high reliability and great controllability at the expense of switching noise and EMI issues. To implement a digitally controlled PWM with high resolution and low EMI, this study proposes a random carrier frequency modulation with dithered duty cycle method, where wide spectrum dithering and random modulation of spread spectrum PWM are tightly integrated. The proposed system was not only able to suppress voltage ripples, but also possible to minimize the occurrence of dependent frequencies due to the fixed dithering pattern. The spectrum of the output voltage showed that the switching frequency component was reduced by about 6.1dB and the output voltage ripples was also reduced compared to the existing PWM techniques.
- Open Access
- 18 February 2021
- Published by MDPI AG in Electronics
- Vol. 10 (4) , 479
Efficient Depth Map Creation with a Lightweight Deep Neural Network
Finding depth information with stereo matching using a deep learning algorithm for embedded systems has recently gained significant attention owing to emerging high-performance mobile graphics processing units (GPUs). Several researchers have proposed feasible small-scale CNNs that can run on a local GPU, but they still suffer from low accuracy and/or high computational requirements. In the method proposed in this study, pooling layers with padding and an asymmetric convolution filter are used to reduce computational costs and simultaneously maintain the accuracy of disparity. The patch size and number of layers are adjusted by analyzing the feature and activation maps. The proposed method forms a small-scale network algorithm suitable for a vision system at the edge and still exhibits high-disparity accuracy and low computational loads as compared to existing stereo-matching networks.
- 18 February 2021
- Published by MDPI AG in Electronics
- Vol. 10 (4) , 479
Open AccessEfficient Depth Map Creation with a Lightweight Deep Neural Network
Finding depth information with stereo matching using a deep learning algorithm for embedded systems has recently gained significant attention owing to emerging high-performance mobile graphics processing units (GPUs). Several researchers have proposed feasible small-scale CNNs that can run on a local GPU, but they still suffer from low accuracy and/or high computational requirements. In the method proposed in this study, pooling layers with padding and an asymmetric convolution filter are used to reduce computational costs and simultaneously maintain the accuracy of disparity. The patch size and number of layers are adjusted by analyzing the feature and activation maps. The proposed method forms a small-scale network algorithm suitable for a vision system at the edge and still exhibits high-disparity accuracy and low computational loads as compared to existing stereo-matching networks.
- 31 October 2019
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 8 (5) , 347-357
An Error Resilience and Concealment Method for Line-based Wavelet Image Compressions
Recently, some wavelet-based line compression methods for low-cost high-resolution image transmission have been studied. With these line compression methods, which are a combination of the multilevel discrete wavelet transform, predictive coding, and entropy coding, communications errors on the transmission channel can cause an image quality–degradation phenomenon of various line shapes. In this paper, we analyze image degradation patterns of wavelet-based hybrid line compression techniques caused by communications errors. Then, we propose an image quality–improvement method including a low-cost resynchronization technique by using a periodic reset to prevent the error propagation phenomenon of predictive coding, as well as an error concealment technique that uses only one additional line memory in the variable reference line–based decoder to improve degraded image quality. Experimental results showed that the proposed method achieved better image quality than the conventional techniques at the small expense of compression ratio while satisfying low-cost design conditions pursued by existing line compression systems.
- 31 October 2019
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 8 (5) , 347-357
An Error Resilience and Concealment Method for Line-based Wavelet Image Compressions
Recently, some wavelet-based line compression methods for low-cost high-resolution image transmission have been studied. With these line compression methods, which are a combination of the multilevel discrete wavelet transform, predictive coding, and entropy coding, communications errors on the transmission channel can cause an image quality–degradation phenomenon of various line shapes. In this paper, we analyze image degradation patterns of wavelet-based hybrid line compression techniques caused by communications errors. Then, we propose an image quality–improvement method including a low-cost resynchronization technique by using a periodic reset to prevent the error propagation phenomenon of predictive coding, as well as an error concealment technique that uses only one additional line memory in the variable reference line–based decoder to improve degraded image quality. Experimental results showed that the proposed method achieved better image quality than the conventional techniques at the small expense of compression ratio while satisfying low-cost design conditions pursued by existing line compression systems.
- Open Access
- 17 September 2018
- Published by Springer Nature in EURASIP Journal on Image and Video Processing
- Vol. 2018 (1) , 87
An energy-efficient low-memory image compression system for multimedia IoT products
- 17 September 2018
- Published by Springer Nature in EURASIP Journal on Image and Video Processing
- Vol. 2018 (1) , 87
Open AccessAn energy-efficient low-memory image compression system for multimedia IoT products
- 30 June 2018
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 7 (3) , 201-209
Depth Estimation with Manhattan World Cues on a Monocular Image
Many technologies requiring three-dimensional (3D) information, such as augmented reality, virtual reality, robots, and autonomous vehicles, are developing rapidly. Among them, 3D reconstruction from a monocular image is challenging because of the lack of information in monocular images and the large-scale errors from wide, low-texture regions found in them. However, a modern image obtained outdoors contains many artificial environmental stereo cues; hence, we regard this as a Manhattan world. In this study, we find vanishing points, which are effective stereo cues in the Manhattan world, to reconstruct 3D information from a monocular image. Then, 3D geometric information in this image is derived based on the vanishing points and relative positions and directions of straight lines. Hence, an initial depth map is generated accordingly. Depending on the 3D geometric information of each stereo cue found, accurate depth values are estimated and assigned to depth maps. As a result, the accuracy of depth information from an outdoor image obtained with the proposed method is improved by 37%, compared to existing algorithms.
- 30 June 2018
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 7 (3) , 201-209
Depth Estimation with Manhattan World Cues on a Monocular Image
Many technologies requiring three-dimensional (3D) information, such as augmented reality, virtual reality, robots, and autonomous vehicles, are developing rapidly. Among them, 3D reconstruction from a monocular image is challenging because of the lack of information in monocular images and the large-scale errors from wide, low-texture regions found in them. However, a modern image obtained outdoors contains many artificial environmental stereo cues; hence, we regard this as a Manhattan world. In this study, we find vanishing points, which are effective stereo cues in the Manhattan world, to reconstruct 3D information from a monocular image. Then, 3D geometric information in this image is derived based on the vanishing points and relative positions and directions of straight lines. Hence, an initial depth map is generated accordingly. Depending on the 3D geometric information of each stereo cue found, accurate depth values are estimated and assigned to depth maps. As a result, the accuracy of depth information from an outdoor image obtained with the proposed method is improved by 37%, compared to existing algorithms.
- 30 October 2016
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 5 (5) , 310-318
A Consistent Quality Bit Rate Control for the Line-Based Compression
- 30 October 2016
- Published by The Institute of Electronics Engineers of Korea in IEIE Transactions on Smart Processing & Computing
- Vol. 5 (5) , 310-318
A Consistent Quality Bit Rate Control for the Line-Based Compression
- Open Access
- 30 September 2016
- Published by Springer Nature in EURASIP Journal on Image and Video Processing
- Vol. 2016 (1) , 29
A parallel camera image signal processor for SIMD architecture
- 30 September 2016
- Published by Springer Nature in EURASIP Journal on Image and Video Processing
- Vol. 2016 (1) , 29
Open AccessA parallel camera image signal processor for SIMD architecture
- 5 March 2015
- Published by Springer Nature
- p. 67-75
ASiPEC: An Application Specific Instruction-Set Processor for High Performance Entropy Coding
- 5 March 2015
- Published by Springer Nature
- p. 67-75
ASiPEC: An Application Specific Instruction-Set Processor for High Performance Entropy Coding