Taewoo Kim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Taewoo Kim has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Chan-Hyun Youn
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
6 shared publications - Changha Lee
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
5 shared publications - Minsu Jeon
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
5 shared publications - Seong-Hwan Kim
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
3 shared publications - Joo-Young Kim
KAIST
2 shared publications - Junsoo Kim
School of Electrical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
2 shared publications
Latest Publications
- 23 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (3) , 830-841
EPU: An Energy-Efficient Explainable AI Accelerator With Sparsity-Free Computation and Heat Map Compression/Pruning
Deep neural networks (DNNs) have recently gained significant prominence in various real-world applications such as image recognition, natural language processing, and autonomous vehicles. However, due to their black-box nature in system, the underlying mechanisms of DNNs behind the inference results remain opaque to users. In order to address this challenge, researchers have focused on developing explainable artificial intelligence (AI) algorithms. Explainable AI aims to provide a clear and human-understandable explanation of the model’s decision, thereby building more reliable systems. However, the explanation task differs from well-known inference and training processes as it involves interactions with the user. Consequently, existing inference and training accelerators face inefficiencies when processing explainable AI on edge devices. This article introduces explainable processing unit (EPU), the first hardware accelerator designed for explainable AI workloads. The EPU utilizes a novel data compression format for the output heat maps and intermediate gradients to enhance the overall system performance by reducing both memory footprint and external memory access. Its sparsity-free computing core efficiently handles the input sparsity with negligible control overhead, resulting in a throughput boost of up to 9.48× It also proposes a dynamic workload scheduling with a customized ON-chip network for distinct inference and explanation tasks to maximize internal data reuse hence reducing external memory access by 63.7%. Furthermore, the EPU incorporates point-wise gradient pruning (PGP) that can significantly reduce the size of heat maps by a factor of 7.01× combined with the proposed compression format. Finally, the EPU chip fabricated in a 28 nm CMOS process achieves a remarkable heat map generation rate of 367 frames/s for ResNet-34 while maintaining the state-of-the-art area and energy efficiency of 112.3 GOPS/mm2 and 26.55 TOPS/W, respectively.
- 23 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (3) , 830-841
EPU: An Energy-Efficient Explainable AI Accelerator With Sparsity-Free Computation and Heat Map Compression/Pruning
Deep neural networks (DNNs) have recently gained significant prominence in various real-world applications such as image recognition, natural language processing, and autonomous vehicles. However, due to their black-box nature in system, the underlying mechanisms of DNNs behind the inference results remain opaque to users. In order to address this challenge, researchers have focused on developing explainable artificial intelligence (AI) algorithms. Explainable AI aims to provide a clear and human-understandable explanation of the model’s decision, thereby building more reliable systems. However, the explanation task differs from well-known inference and training processes as it involves interactions with the user. Consequently, existing inference and training accelerators face inefficiencies when processing explainable AI on edge devices. This article introduces explainable processing unit (EPU), the first hardware accelerator designed for explainable AI workloads. The EPU utilizes a novel data compression format for the output heat maps and intermediate gradients to enhance the overall system performance by reducing both memory footprint and external memory access. Its sparsity-free computing core efficiently handles the input sparsity with negligible control overhead, resulting in a throughput boost of up to 9.48× It also proposes a dynamic workload scheduling with a customized ON-chip network for distinct inference and explanation tasks to maximize internal data reuse hence reducing external memory access by 63.7%. Furthermore, the EPU incorporates point-wise gradient pruning (PGP) that can significantly reduce the size of heat maps by a factor of 7.01× combined with the proposed compression format. Finally, the EPU chip fabricated in a 28 nm CMOS process achieves a remarkable heat map generation rate of 367 frames/s for ResNet-34 while maintaining the state-of-the-art area and energy efficiency of 112.3 GOPS/mm2 and 26.55 TOPS/W, respectively.
- 4 January 2024
- Published by Springer Nature
- p. 19-29
SLO-Aware DL Job Scheduling for Efficient FPGA-GPU Edge Cloud Computing
- 4 January 2024
- Published by Springer Nature
- p. 19-29
SLO-Aware DL Job Scheduling for Efficient FPGA-GPU Edge Cloud Computing
- Open Access
- 9 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 4306-4321
Recursive Visual Explanations Mediation Scheme Based on DropAttention Model With Multiple Episodes Pool
In some DL applications such as remote sensing, it is hard to obtain the high task performance ( e.g . accuracy) using the DL model on image analysis due to the low resolution characteristics of the imagery. Accordingly, several studies attempted to provide visual explanations or apply the attention mechanism to enhance the reliability on the image analysis. However, there still remains structural complexity on obtaining a sophisticated visual explanation with such existing methods: 1) which layer will the visual explanation be extracted from, and 2) which layers the attention modules will be applied to. 3) Subsequently, in order to observe the aspects of visual explanations on such diverse episodes of applying attention modules individually, training cost inefficiency inevitably arises as it requires training the multiple models one by one in the conventional methods. In order to solve the problems, we propose a new scheme of mediating the visual explanations in a pixel-level recursively. Specifically, we propose DropAtt that generates multiple episodes pool by training only a single network once as an amortized model, which also shows stability on task performance regardless of layer-wise attention policy. From the multiple episodes pool generated by DropAtt, by quantitatively evaluating the explainability of each visual explanation and expanding the parts of explanations with high explainability recursively, our visual explanations mediatio scheme attempts to adjust how much to reflect each episodic layer-wise explanation for enforcing a dominant explainability of each candidate. On the empirical evaluation, our methods show their feasibility on enhancing the visual explainability by reducing average drop about 17% and enhancing the rate of increase in confidence 3%.
- 9 January 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 4306-4321
Open AccessRecursive Visual Explanations Mediation Scheme Based on DropAttention Model With Multiple Episodes Pool
In some DL applications such as remote sensing, it is hard to obtain the high task performance ( e.g . accuracy) using the DL model on image analysis due to the low resolution characteristics of the imagery. Accordingly, several studies attempted to provide visual explanations or apply the attention mechanism to enhance the reliability on the image analysis. However, there still remains structural complexity on obtaining a sophisticated visual explanation with such existing methods: 1) which layer will the visual explanation be extracted from, and 2) which layers the attention modules will be applied to. 3) Subsequently, in order to observe the aspects of visual explanations on such diverse episodes of applying attention modules individually, training cost inefficiency inevitably arises as it requires training the multiple models one by one in the conventional methods. In order to solve the problems, we propose a new scheme of mediating the visual explanations in a pixel-level recursively. Specifically, we propose DropAtt that generates multiple episodes pool by training only a single network once as an amortized model, which also shows stability on task performance regardless of layer-wise attention policy. From the multiple episodes pool generated by DropAtt, by quantitatively evaluating the explainability of each visual explanation and expanding the parts of explanations with high explainability recursively, our visual explanations mediatio scheme attempts to adjust how much to reflect each episodic layer-wise explanation for enforcing a dominant explainability of each candidate. On the empirical evaluation, our methods show their feasibility on enhancing the visual explainability by reducing average drop about 17% and enhancing the rate of increase in confidence 3%.
- Open Access
- 27 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 7043-7055
A Channel Pruning Optimization With Layer-Wise Sensitivity in a Single-Shot Manner Under Computational Constraints
In the constrained computing environments such as mobile device or satellite on-board system, various computational factors of hardware resource can restrict the processing of deep learning (DL) services. Recent DL models such as satellite image analysis mainly require larger resource memory occupation for intermediate feature map footprint than the given memory specification of hardware resource and larger computational overhead (in FLOP) to meet service-level objective in the sense of hardware accelerator. As one of the solutions, we propose a new method of controlling the layer-wise channel pruning in a single-shot manner that can decide how much channels to prune in each layer by observing dataset once without full pretraining. To improve the robustness of the performance degradation, we also propose a layer-wise sensitivity and formulate the optimization problems for deciding layer-wise pruning ratio under target computational constraints. In the paper, the optimal conditions are theoretically derived, and the practical optimum searching schemes are proposed using the optimal conditions. On the empirical evaluation, the proposed methods show robustness on performance degradation, and present feasibility on DL serving under constrained computing environments by reducing memory occupation, providing acceleration effect and throughput improvement while keeping the accuracy performance.
- 27 December 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 7043-7055
Open AccessA Channel Pruning Optimization With Layer-Wise Sensitivity in a Single-Shot Manner Under Computational Constraints
In the constrained computing environments such as mobile device or satellite on-board system, various computational factors of hardware resource can restrict the processing of deep learning (DL) services. Recent DL models such as satellite image analysis mainly require larger resource memory occupation for intermediate feature map footprint than the given memory specification of hardware resource and larger computational overhead (in FLOP) to meet service-level objective in the sense of hardware accelerator. As one of the solutions, we propose a new method of controlling the layer-wise channel pruning in a single-shot manner that can decide how much channels to prune in each layer by observing dataset once without full pretraining. To improve the robustness of the performance degradation, we also propose a layer-wise sensitivity and formulate the optimization problems for deciding layer-wise pruning ratio under target computational constraints. In the paper, the optimal conditions are theoretically derived, and the practical optimum searching schemes are proposed using the optimal conditions. On the empirical evaluation, the proposed methods show robustness on performance degradation, and present feasibility on DL serving under constrained computing environments by reducing memory occupation, providing acceleration effect and throughput improvement while keeping the accuracy performance.
- Open Access
- 4 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 117315-117333
Federated Onboard-Ground Station Computing With Weakly Supervised Cascading Pyramid Attention Network for Satellite Image Analysis
With advances in NanoSat (CubeSat) and high-resolution sensors, the amount of raw data to be analyzed by human supervisors has been explosively increasing for satellite image analysis. To reduce the raw data, the satellite onboard AI processing with low-power COTS (Commercial, Off-The-Shelf) HW has emerged from a real satellite mission. It filters the useless data (e.g. cloudy images) that is worthless to supervisors, achieving efficient satellite-ground station communication. In the application for complex object recognition, however, additional explanation is required for the reliability of the AI prediction due to its low performance. Although various eXplainable AI (XAI) methods for providing human-interpretable explanation have been studied, the pyramid architecture in a deep network leads to the background bias problem which visual explanation only focuses on the background context around the object. Missing the small objects in a tiny region leads to poor explainability although the AI model corrects the object class. To resolve the problems, we propose a novel federated onboard-ground station (FOGS) computing with Cascading Pyramid Attention Network (CPANet) for reliable onboard XAI in object recognition. We present an XAI architecture with a cascading attention mechanism for mitigating the background bias for the onboard processing. By exploiting the localization ability in pyramid feature blocks, we can extract high-quality visual explanation covering the both semantic and small contexts of an object. For enhancing visual explainability of complex satellite images, we also describe a novel computing federation with the ground station and supervisors. In the ground station, active learning-based sample selection and attention refinement scheme with a simple feedback method are conducted to achieve the robustness of explanation and efficient supervisor’s annotation cost, simultaneously. Experiments on various datasets show that the proposed system improves accuracy in object recognition and accurate visual explanation detecting small contexts of objects even in a peripheral region. Then, our attention refinement mechanism demonstrates that the inconsistent explanation can be efficiently resolved only with very simple selection-based feedback.
- 4 November 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 117315-117333
Open AccessFederated Onboard-Ground Station Computing With Weakly Supervised Cascading Pyramid Attention Network for Satellite Image Analysis
With advances in NanoSat (CubeSat) and high-resolution sensors, the amount of raw data to be analyzed by human supervisors has been explosively increasing for satellite image analysis. To reduce the raw data, the satellite onboard AI processing with low-power COTS (Commercial, Off-The-Shelf) HW has emerged from a real satellite mission. It filters the useless data (e.g. cloudy images) that is worthless to supervisors, achieving efficient satellite-ground station communication. In the application for complex object recognition, however, additional explanation is required for the reliability of the AI prediction due to its low performance. Although various eXplainable AI (XAI) methods for providing human-interpretable explanation have been studied, the pyramid architecture in a deep network leads to the background bias problem which visual explanation only focuses on the background context around the object. Missing the small objects in a tiny region leads to poor explainability although the AI model corrects the object class. To resolve the problems, we propose a novel federated onboard-ground station (FOGS) computing with Cascading Pyramid Attention Network (CPANet) for reliable onboard XAI in object recognition. We present an XAI architecture with a cascading attention mechanism for mitigating the background bias for the onboard processing. By exploiting the localization ability in pyramid feature blocks, we can extract high-quality visual explanation covering the both semantic and small contexts of an object. For enhancing visual explainability of complex satellite images, we also describe a novel computing federation with the ground station and supervisors. In the ground station, active learning-based sample selection and attention refinement scheme with a simple feedback method are conducted to achieve the robustness of explanation and efficient supervisor’s annotation cost, simultaneously. Experiments on various datasets show that the proposed system improves accuracy in object recognition and accurate visual explanation detecting small contexts of objects even in a peripheral region. Then, our attention refinement mechanism demonstrates that the inconsistent explanation can be efficiently resolved only with very simple selection-based feedback.
- Open Access
- 30 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 37164-37180
FleX: A Flex Interconnected HPC System With Stochastic Load Balancing Scheme
We propose a new low-diameter interconnection network called FleX, which offers high flexibility when installing interconnections in a HPC system. FleX consists of multiple layers with only connections between neighboring layers and not within each layer. These structural properties make it easy to achieve a low diameter with regardless of the scale. The cross-like connections between the adjacent layers in FleX impart various alternative minimal paths, allowing FleX to have high resiliency and a wide bisection width. We also discuss the minimal routing scheme and a stochastic load balancing scheme (LBR) for the proposed interconnection network. Through cycle-based simulations, the performance of FleX is evaluated, and the cost and power consumption analyses in comparison with other interconnection networks are also conducted. We verify that FleX has high configuration flexibility with regard to cost and performance, and also provides low latency and high saturation throughput with the same cost over the legacy interconnection networks for the HPC system. Moreover, being synergied with the proposing LBR, we also verify that FleX can expand its saturation throughput further while only sacrificing the latency slightly.
- 30 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 37164-37180
Open AccessFleX: A Flex Interconnected HPC System With Stochastic Load Balancing Scheme
We propose a new low-diameter interconnection network called FleX, which offers high flexibility when installing interconnections in a HPC system. FleX consists of multiple layers with only connections between neighboring layers and not within each layer. These structural properties make it easy to achieve a low diameter with regardless of the scale. The cross-like connections between the adjacent layers in FleX impart various alternative minimal paths, allowing FleX to have high resiliency and a wide bisection width. We also discuss the minimal routing scheme and a stochastic load balancing scheme (LBR) for the proposed interconnection network. Through cycle-based simulations, the performance of FleX is evaluated, and the cost and power consumption analyses in comparison with other interconnection networks are also conducted. We verify that FleX has high configuration flexibility with regard to cost and performance, and also provides low latency and high saturation throughput with the same cost over the legacy interconnection networks for the HPC system. Moreover, being synergied with the proposing LBR, we also verify that FleX can expand its saturation throughput further while only sacrificing the latency slightly.
- 1 May 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1921-1926
Hybrid DC-DC Converter Using Center-Tapped Clamp Circuit in Wide Range of High Output Voltage
In this paper, a hybrid converter with a new clamp circuit is presented. This hybrid converter has the structure of combining a phase-shifted full-bridge (PSFB)converter with a half-bridge (HB)LLC converter. On the primary side, the lagging-leg of PSFB converter is shared by HB LLC converter. On the secondary side, the full-bridge rectifiers of two converters are connected in series. Due to the series-connected structure on the secondary side, this hybrid converter is suitable for high output voltage applications. In addition, by using a new center-tapped clamp circuit, which consists of two diodes and two capacitors with connected to the center of the transformer, the proposed converter can solve many drawbacks of PSFB converters in wide range of output voltage. Finally, the performance in efficiency, power density, and cost can be improved. In order to validate the feasibility of the proposed converter, a 3.3-kW prototype was tested.
- 1 May 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1921-1926
Hybrid DC-DC Converter Using Center-Tapped Clamp Circuit in Wide Range of High Output Voltage
In this paper, a hybrid converter with a new clamp circuit is presented. This hybrid converter has the structure of combining a phase-shifted full-bridge (PSFB)converter with a half-bridge (HB)LLC converter. On the primary side, the lagging-leg of PSFB converter is shared by HB LLC converter. On the secondary side, the full-bridge rectifiers of two converters are connected in series. Due to the series-connected structure on the secondary side, this hybrid converter is suitable for high output voltage applications. In addition, by using a new center-tapped clamp circuit, which consists of two diodes and two capacitors with connected to the center of the transformer, the proposed converter can solve many drawbacks of PSFB converters in wide range of output voltage. Finally, the performance in efficiency, power density, and cost can be improved. In order to validate the feasibility of the proposed converter, a 3.3-kW prototype was tested.