Luhong Liang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Luhong Liang has a total of 55 co-authors. Here are the top co-authors by number of shared publications:
- Kwang-Ting Cheng
Department of Electronic and Computer Engineering, the Hong Kong University of Science and Technology, Hong Kong, China
6 shared publications - Yu Liu
Hong Kong University of Science and Technology
4 shared publications - Peng Luo
Hong Kong University of Science and Technology
3 shared publications - Pingcheng Dong
Hong Kong University of Science and Technology
3 shared publications Xuejiao Liu
Hong Kong University of Science and Technology
3 shared publications- Yonghao Tan
Hong Kong University of Science and Technology
3 shared publications
Latest Publications
- 2 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. PP (02780070) , 1
Towards Secure and Efficient DNN Accelerators via Hardware-Software Co-Design
The rapid deployment of deep neural network (DNN) accelerators in safety-critical domains such as autonomous vehicles, healthcare systems, and financial infrastructure necessitates robust mechanisms to safeguard data confidentiality and computational integrity. Existing security solutions for DNN accelerators, however, suffer from excessive hardware resource demands and frequent off-chip memory access overheads, which degrade performance and scalability. To address these challenges, this paper presents a secure and efficient memory protection framework for DNN accelerators with minimal overhead. First, we propose a bandwidth-aware cryptographic scheme that adapts encryption granularity based on memory traffic patterns, striking a balance between security and resource efficiency. Second, we observe that both the overlapping regions in the intra-layer tiling’s sliding window pattern and those resulting from inter-layer tiling strategy discrepancies introduce substantial redundant memory accesses and repeated computational overhead in cryptography. Third, we introduce a multi-level authentication mechanism that effectively eliminates unnecessary off-chip memory accesses, enhancing performance and energy efficiency. Experimental results show that this work decreases performance overhead by over 12% and achieves 87% energy efficiency improvement for both server and edge neural processing units (NPUs), while ensuring robust scalability.
- 2 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. PP (02780070) , 1
Towards Secure and Efficient DNN Accelerators via Hardware-Software Co-Design
The rapid deployment of deep neural network (DNN) accelerators in safety-critical domains such as autonomous vehicles, healthcare systems, and financial infrastructure necessitates robust mechanisms to safeguard data confidentiality and computational integrity. Existing security solutions for DNN accelerators, however, suffer from excessive hardware resource demands and frequent off-chip memory access overheads, which degrade performance and scalability. To address these challenges, this paper presents a secure and efficient memory protection framework for DNN accelerators with minimal overhead. First, we propose a bandwidth-aware cryptographic scheme that adapts encryption granularity based on memory traffic patterns, striking a balance between security and resource efficiency. Second, we observe that both the overlapping regions in the intra-layer tiling’s sliding window pattern and those resulting from inter-layer tiling strategy discrepancies introduce substantial redundant memory accesses and repeated computational overhead in cryptography. Third, we introduce a multi-level authentication mechanism that effectively eliminates unnecessary off-chip memory accesses, enhancing performance and energy efficiency. Experimental results show that this work decreases performance overhead by over 12% and achieves 87% energy efficiency improvement for both server and edge neural processing units (NPUs), while ensuring robust scalability.
- 15 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (4) , 2647-2658
An Adaptive Depth Processing System Based on Foundation Transformers and Time-of-Flight Fusion
Generating high-quality depth maps with accurate values is a critical research topic, and various depth estimation methods, such as Time-of-Flight (ToF) and Monocular Depth Estimation (MDE), are advancing rapidly. However, each single-sensor approach is inherently limited by the characteristics of its respective modality. Meanwhile, fused approaches using multiple sensors or dedicated trained models are often plagued by system complexity and limited generalization. In this paper, we propose an adaptive depth processing system based on the foundation transformer and ToF fusion, aiming to harness the precision of ToF data and the high-quality depth priors provided by the foundation model in a monocular and zero-shot manner. A hardware–algorithm co-design partitions the computation between a dedicated ASIC for compute-intensive foundation transformer acceleration and an MPSoC FPGA for potential reconfigurable fusion, yielding 36.4 ms latency and 1.9 W power under a 27.6 GOPS/frame workload. Zero-shot evaluations on various ToF data sources including FLAT, TICaM, and in-house captures confirm strong cross-domain generalization, demonstrating the system’s adaptability, efficiency, and accuracy.
- 15 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (4) , 2647-2658
An Adaptive Depth Processing System Based on Foundation Transformers and Time-of-Flight Fusion
Generating high-quality depth maps with accurate values is a critical research topic, and various depth estimation methods, such as Time-of-Flight (ToF) and Monocular Depth Estimation (MDE), are advancing rapidly. However, each single-sensor approach is inherently limited by the characteristics of its respective modality. Meanwhile, fused approaches using multiple sensors or dedicated trained models are often plagued by system complexity and limited generalization. In this paper, we propose an adaptive depth processing system based on the foundation transformer and ToF fusion, aiming to harness the precision of ToF data and the high-quality depth priors provided by the foundation model in a monocular and zero-shot manner. A hardware–algorithm co-design partitions the computation between a dedicated ASIC for compute-intensive foundation transformer acceleration and an MPSoC FPGA for potential reconfigurable fusion, yielding 36.4 ms latency and 1.9 W power under a 27.6 GOPS/frame workload. Zero-shot evaluations on various ToF data sources including FLAT, TICaM, and in-house captures confirm strong cross-domain generalization, demonstrating the system’s adaptability, efficiency, and accuracy.
- Open Access
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 315-316
A 10.60 μW 150 GOPS Mixed-Bit-Width Sparse CNN Accelerator for Life-Threatening Ventricular Arrhythmia Detection
This paper proposes an ultra-low power, mixed-bit-width sparse convolutional neural network (CNN) accelerator to accelerate ventricular arrhythmia (VA) detection. The chip achieves 50% sparsity in a quantized 1D CNN using a sparse processing element (SPE) architecture. Measurement on the prototype chip TSMC 40nm CMOS low-power (LP) process for the VA classification task demonstrates that it consumes 10.60 μW of power while achieving a performance of 150 GOPS and a diagnostic accuracy of 99.95%. The computation power density is only 0.57 μW/mm2, which is 14.23× smaller than state-of-the-art works, making it highly suitable for implantable and wearable medical devices.
- 20 January 2025
- Published by Association for Computing Machinery (ACM)
- p. 315-316
Open AccessA 10.60 μW 150 GOPS Mixed-Bit-Width Sparse CNN Accelerator for Life-Threatening Ventricular Arrhythmia Detection
This paper proposes an ultra-low power, mixed-bit-width sparse convolutional neural network (CNN) accelerator to accelerate ventricular arrhythmia (VA) detection. The chip achieves 50% sparsity in a quantized 1D CNN using a sparse processing element (SPE) architecture. Measurement on the prototype chip TSMC 40nm CMOS low-power (LP) process for the VA classification task demonstrates that it consumes 10.60 μW of power while achieving a performance of 150 GOPS and a diagnostic accuracy of 99.95%. The computation power density is only 0.57 μW/mm2, which is 14.23× smaller than state-of-the-art works, making it highly suitable for implantable and wearable medical devices.
- Open Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Genetic Quantization-Aware Approximation for Non-Linear Operations in Transformers
Non-linear functions are prevalent in Transformers and their lightweight variants, incurring substantial and frequently underestimated hardware costs. Previous state-of-the-art works optimize these operations by piece-wise linear approximation and store the parameters in look-up tables (LUT), but most of them require unfriendly high-precision arithmetics such as FP/INT 32 and lack consideration of integer-only INT quantization. This paper proposed a genetic LUT-Approximation algorithm namely GQA-LUT that can automatically determine the parameters with quantization awareness. The results demonstrate that GQA-LUT achieves negligible degradation on the challenging semantic segmentation task for both vanilla and linear Transformer models. Besides, proposed GQA-LUT enables the employment of INT8-based LUT-Approximation that achieves an area savings of 81.3~81.7% and a power reduction of 79.3~80.2% compared to the high-precision FP/INT 32 alternatives. Code is available at https://github.com/PingchengDong/GQA-LUT.
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessGenetic Quantization-Aware Approximation for Non-Linear Operations in Transformers
Non-linear functions are prevalent in Transformers and their lightweight variants, incurring substantial and frequently underestimated hardware costs. Previous state-of-the-art works optimize these operations by piece-wise linear approximation and store the parameters in look-up tables (LUT), but most of them require unfriendly high-precision arithmetics such as FP/INT 32 and lack consideration of integer-only INT quantization. This paper proposed a genetic LUT-Approximation algorithm namely GQA-LUT that can automatically determine the parameters with quantization awareness. The results demonstrate that GQA-LUT achieves negligible degradation on the challenging semantic segmentation task for both vanilla and linear Transformer models. Besides, proposed GQA-LUT enables the employment of INT8-based LUT-Approximation that achieves an area savings of 81.3~81.7% and a power reduction of 79.3~80.2% compared to the high-precision FP/INT 32 alternatives. Code is available at https://github.com/PingchengDong/GQA-LUT.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (8) , 1400-1414
WASP: Efficient Power Management Enabling Workload-Aware, Self-Powered AIoT Devices
The wide adoption of edge AI has heightened the demand for various battery-less and maintenance-free smart systems. Nevertheless, emerging Artificial Intelligence of Things (AIoT) are complex workloads showing increased power demand, diversified power usage patterns, and unique sensitivity to power management (PM) approaches. Existing AIoT devices cannot select the most appropriate PM tuning knob, and therefore they often make sub-optimal decisions. In addition, these PM solutions always assume traditional power regulation circuit which incurs non-negligible power loss and control overhead. This can greatly compromise the potential of AIoT efficiency. In this paper, we explore power management optimization for emerging self-powered AIoT devices. We propose WASP, a highly efficient power management scheme for workload-aware, self-powered AIoT devices. The novelty of WASP is two fold. First, it combines offline profiling and light-weight online control to select the most appropriate PM tuning knobs for the given DNN models. Second, it is well tailored to a reconfigurable voltage regulation module that can make the best use of the limited power budget. Our results show that WASP allows AIoT devices to accomplish 65.6% more inference tasks under a stringent power budget without any performance degradation compared with other existing approaches.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 35 (8) , 1400-1414
WASP: Efficient Power Management Enabling Workload-Aware, Self-Powered AIoT Devices
The wide adoption of edge AI has heightened the demand for various battery-less and maintenance-free smart systems. Nevertheless, emerging Artificial Intelligence of Things (AIoT) are complex workloads showing increased power demand, diversified power usage patterns, and unique sensitivity to power management (PM) approaches. Existing AIoT devices cannot select the most appropriate PM tuning knob, and therefore they often make sub-optimal decisions. In addition, these PM solutions always assume traditional power regulation circuit which incurs non-negligible power loss and control overhead. This can greatly compromise the potential of AIoT efficiency. In this paper, we explore power management optimization for emerging self-powered AIoT devices. We propose WASP, a highly efficient power management scheme for workload-aware, self-powered AIoT devices. The novelty of WASP is two fold. First, it combines offline profiling and light-weight online control to select the most appropriate PM tuning knobs for the given DNN models. Second, it is well tailored to a reconfigurable voltage regulation module that can make the best use of the limited power budget. Our results show that WASP allows AIoT devices to accomplish 65.6% more inference tasks under a stringent power budget without any performance degradation compared with other existing approaches.
- Open Access
- 17 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-13
Architecting Efficient Multi-modal AIoT Systems
Multi-modal computing (M2C) has recently exhibited impressive accuracy improvements in numerous autonomous artificial intelligence of things (AIoT) systems. However, this accuracy gain is often tethered to an incredible increase in energy consumption. Particularly, various highly-developed modality sensors devour most of the energy budget, which would make the deployment of M2C for real-world AIoT applications a difficult challenge. To address the above issue, we propose AMG, an innovative HW/SW co-design solution tailored to multi-modal AIoT systems. The key behind AMG is modality gating (throttling) that allows for adaptively sensing and computing modalities for different tasks. This is non-trivial since we must balance situational awareness, energy conservation, and execution latency. AMG achieves our goal with two first-of-its-kind designs. 1) It introduces a novel decoupled modality sensor architecture to support partial throttling of modality sensors. Doing so allows one to greatly save AIoT power but maintains sensor data flow. 2) AMG also features a smart power management strategy based on the new architecture, allowing the device to initialize and tune itself with the optimal configuration. It can predict whether a reasonable degree of accuracy will be satisfied during runtime, and react proactively to remediate the gating process. Extensive evaluation based on our prototype system confirms that AMG improves the AIoT lifespan by 74.5% to 133.7% with the same energy budget while meeting the performance requirements.
- 17 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-13
Open AccessArchitecting Efficient Multi-modal AIoT Systems
Multi-modal computing (M2C) has recently exhibited impressive accuracy improvements in numerous autonomous artificial intelligence of things (AIoT) systems. However, this accuracy gain is often tethered to an incredible increase in energy consumption. Particularly, various highly-developed modality sensors devour most of the energy budget, which would make the deployment of M2C for real-world AIoT applications a difficult challenge. To address the above issue, we propose AMG, an innovative HW/SW co-design solution tailored to multi-modal AIoT systems. The key behind AMG is modality gating (throttling) that allows for adaptively sensing and computing modalities for different tasks. This is non-trivial since we must balance situational awareness, energy conservation, and execution latency. AMG achieves our goal with two first-of-its-kind designs. 1) It introduces a novel decoupled modality sensor architecture to support partial throttling of modality sensors. Doing so allows one to greatly save AIoT power but maintains sensor data flow. 2) AMG also features a smart power management strategy based on the new architecture, allowing the device to initialize and tune itself with the optimal configuration. It can predict whether a reasonable degree of accuracy will be satisfied during runtime, and react proactively to remediate the gating process. Extensive evaluation based on our prototype system confirms that AMG improves the AIoT lifespan by 74.5% to 133.7% with the same energy budget while meeting the performance requirements.
- 15 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 70 (8) , 3079-3083
A Tiny Accelerator for Mixed-Bit Sparse CNN Based on Efficient Fetch Method of SIMO SPad
Convolution neural networks (CNNs) have been implemented with custom hardware on edge devices since its algorithms were successful in many artificial intelligence applications. Although lots of unstructured pruning and mix-bit quantization algorithms have been proposed to successfully compress CNNs, there are few hardware accelerators which can support both sparse and mix-bit CNNs. Besides, sparse matrix computation consumes lots of hardware resources such as registers or BRAM to fetch the needed input activations into processing element (PE). This paper presents a tiny accelerator for mixed-bit sparse CNNs featuring a novel scheme of single vector-based compressed sparse filter (CSF) method and single input multiple output scratch pad (SIMO SPad) to effectively compress weight and fetch the needed input activation. SIMO SPad is shared by multiple PEs, which saves 13.34% CLB LUTs, 46.24% CLB Registers. Furthermore, the accelerator supports mixed-bit sparse computation to obtain better accuracy and performance. When tested on VGG16, compared with 8-bit non-sparsity baselines, the performance of mixed-bit sparsity on Cifar10 and ImageNet improved by 4.85× and 3.33×, respectively, with small accuracy decrease degradation. Compared to state-of-the-art accelerators, the accelerator achieves 1.40× to 2.98× greater DSP efficiency, and offers 1.91× greater energy efficiency.
- 15 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 70 (8) , 3079-3083
A Tiny Accelerator for Mixed-Bit Sparse CNN Based on Efficient Fetch Method of SIMO SPad
Convolution neural networks (CNNs) have been implemented with custom hardware on edge devices since its algorithms were successful in many artificial intelligence applications. Although lots of unstructured pruning and mix-bit quantization algorithms have been proposed to successfully compress CNNs, there are few hardware accelerators which can support both sparse and mix-bit CNNs. Besides, sparse matrix computation consumes lots of hardware resources such as registers or BRAM to fetch the needed input activations into processing element (PE). This paper presents a tiny accelerator for mixed-bit sparse CNNs featuring a novel scheme of single vector-based compressed sparse filter (CSF) method and single input multiple output scratch pad (SIMO SPad) to effectively compress weight and fetch the needed input activation. SIMO SPad is shared by multiple PEs, which saves 13.34% CLB LUTs, 46.24% CLB Registers. Furthermore, the accelerator supports mixed-bit sparse computation to obtain better accuracy and performance. When tested on VGG16, compared with 8-bit non-sparsity baselines, the performance of mixed-bit sparsity on Cifar10 and ImageNet improved by 4.85× and 3.33×, respectively, with small accuracy decrease degradation. Compared to state-of-the-art accelerators, the accelerator achieves 1.40× to 2.98× greater DSP efficiency, and offers 1.91× greater energy efficiency.