Query：
学者姓名：廖龙龙
Refining：
Year
Type
Indexed by
Source
Complex
Former Name
Co-
Language
Clean All
Abstract ：
Deep learning inference on edge devices is susceptible to security threats, particularly Fault Injection Attacks (FIAs), which are easily executed and pose a significant risk to the inference. These attacks could potentially lead to alterations to the memory of the edge device or errors in the execution of instructions. Specifically, time-intensive convolution computation is considerably vulnerable in deep learning inference at the edge. To detect and defend attacks against deep learning inference on heterogeneous edge devices, we propose an efficient hardware-based solution for verifiable model inference named DarkneTV. It leverages an asynchronous mechanism to conduct the hash checking of convolution weights and the verification of convolution computations within the Trusted Execution Environment (TEE) of the Central Processing Unit (CPU) when the integrated Graphics Processing Unit (GPU) runs model inference. It protects the integrity of convolution weights and the correctness of inference results, and effectively detects abnormal weight modifications and incorrect inference results regarding neural operators. Extensive experimental results show that DarkneTV identifies tiny FIAs against convolution weights and computation with over 99.03% accuracy but less extra time overhead. The asynchronous mechanism significantly improves the performance of verifiable inference. Typically, the speedups of the GPU-accelerated verifiable inference on the Hikey 960 achieve 8.50x-11.31x compared with the CPU-only mode. IEEE
Keyword ：
Computational modeling Computational modeling Convolution Convolution Deep learning Deep learning deep learning inference deep learning inference fault injection attacks fault injection attacks Graphics processing units Graphics processing units heterogeneous edge devices heterogeneous edge devices Image edge detection Image edge detection Inference algorithms Inference algorithms Performance evaluation Performance evaluation Trusted Execution Environment Trusted Execution Environment verifiable learning verifiable learning
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Liao, L. , Zheng, Y. , Lu, H. et al. Verifiable Deep Learning Inference on Heterogeneous Edge Devices with Trusted Execution Environment [J]. | IEEE Sensors Journal , 2024 , 24 (17) : 1-1 . |
| MLA | Liao, L. et al. "Verifiable Deep Learning Inference on Heterogeneous Edge Devices with Trusted Execution Environment" . | IEEE Sensors Journal 24 . 17 (2024) : 1-1 . |
| APA | Liao, L. , Zheng, Y. , Lu, H. , Liu, X. , Chen, S. , Yu, Y. . Verifiable Deep Learning Inference on Heterogeneous Edge Devices with Trusted Execution Environment . | IEEE Sensors Journal , 2024 , 24 (17) , 1-1 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Deep learning inference on edge devices is susceptible to security threats, particularly fault injection attacks (FIAs), which are easily executed and pose a significant risk to the inference. These attacks could potentially lead to alterations to the memory of the edge device or errors in the execution of instructions. Specifically, time-intensive convolution computation is considerably vulnerable in deep learning inference at the edge. To detect and defend attacks against deep learning inference on heterogeneous edge devices, we propose an efficient hardware-based solution for verifiable model inference named DarkneTV. It leverages an asynchronous mechanism to conduct hash checking of convolution weights and verification of convolution computations within the trusted execution environment (TEE) of the central processing unit (CPU) when the integrated graphics processing unit (GPU) runs model inference. It protects the integrity of convolution weights and the correctness of inference results, and effectively detects abnormal weight modifications and incorrect inference results regarding neural operators. Extensive experimental results show that DarkneTV identifies tiny FIAs against convolution weights and computation with over 99.03% accuracy but less extra time overhead. The asynchronous mechanism significantly improves the performance of verifiable inference. Typically, the speedups of the GPU-accelerated verifiable inference on the Hikey 960 achieve 8.50 x -11.31 x compared with the CPU-only mode.
Keyword ：
Deep learning inference Deep learning inference fault injection attacks fault injection attacks heterogeneous edge devices heterogeneous edge devices trusted execution environment (TEE) trusted execution environment (TEE) verifiable learning verifiable learning
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Liao, Longlong , Zheng, Yuqiang , Lu, Hong et al. Verifiable Deep Learning Inference on Heterogeneous Edge Devices With Trusted Execution Environment [J]. | IEEE SENSORS JOURNAL , 2024 , 24 (17) : 28351-28362 . |
| MLA | Liao, Longlong et al. "Verifiable Deep Learning Inference on Heterogeneous Edge Devices With Trusted Execution Environment" . | IEEE SENSORS JOURNAL 24 . 17 (2024) : 28351-28362 . |
| APA | Liao, Longlong , Zheng, Yuqiang , Lu, Hong , Liu, Xinqi , Chen, Shuguang , Yu, Yuanlong . Verifiable Deep Learning Inference on Heterogeneous Edge Devices With Trusted Execution Environment . | IEEE SENSORS JOURNAL , 2024 , 24 (17) , 28351-28362 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
CPU-GPU integrated edge devices and deep learning algorithms have received significant progress in recent years, leading to increasingly widespread application of edge intelligence. However, deep learning inference on these edge devices is vulnerable to Fault Injection Attacks (FIAs) that can modify device memory or execute instructions with errors. We propose DarkneTF, a Fault-Tolerant (FT) deep learning inference framework for CPU-GPU integrated edge devices, to ensure the correctness of model inference results by detecting the threat of FIAs. DarkneTF introduces algorithm-based verification to implement the FT deep learning inference. The verification process involves verifying the integrity of model weights and validating the correctness of time- intensive calculations, such as convolutions. We improve the Freivalds algorithm to enhance the ability to detect tiny perturbations by strengthening randomization. As the verification process is also susceptible to FIAs, DarkneTF offloads the verification process into Trusted Execution Environments (TEEs). This scheme ensures the verification process's security and allows for accelerated model inference using the integrated GPUs. Experimental results show that GPU-accelerated FT inference on HiKey 960 achieves notable speedups ranging from 3.46x to 5.57x compared to FT inference on a standalone CPU. The extra memory overhead incurred FT inference remains at an exceedingly low level, with a range of 0.46% to 10.22%. The round-off error of the improved Freivalds algorithm is below 2.50 . 50 x 10 -4 , and the accuracy of detecting FIAs is above 92.73%.
Keyword ：
CPU-GPU integrated edge device CPU-GPU integrated edge device Deep learning Deep learning Fault injection attack Fault injection attack Fault-tolerant inference Fault-tolerant inference Trusted Execution Environment Trusted Execution Environment
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Xu, Hongjian , Liao, Longlong , Liu, Xinqi et al. Fault-tolerant deep learning inference on CPU-GPU integrated edge devices with TEEs [J]. | FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE , 2024 , 161 : 404-414 . |
| MLA | Xu, Hongjian et al. "Fault-tolerant deep learning inference on CPU-GPU integrated edge devices with TEEs" . | FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 161 (2024) : 404-414 . |
| APA | Xu, Hongjian , Liao, Longlong , Liu, Xinqi , Chen, Shuguang , Chen, Jianguo , Liang, Zhixuan et al. Fault-tolerant deep learning inference on CPU-GPU integrated edge devices with TEEs . | FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE , 2024 , 161 , 404-414 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
为实现实验室科研管理过程中的成员工时和工效分析、任务分配的合理性评估等需求,研究一种基于摄像头视频、考勤机记录、Web系统记录等的多模态工效分析系统MASRE.该系统通过实验室科研人员工时及其玩手机行为导致的无效工时、工效实时对比与展示,激励实验室成员投入更多的时间开展学术研究.依据系统计算的工效变化趋势,实验室负责人可分析科研任务分配的合理性,科研人员也可分析影响其科研效率的因素.MASRE系统由负责工时工效统计的Web系统模块和支持无效工时自动识别的AI分析模块构成,采用PyTorch、VUE 3、MySQL等技术实现.以该系统研发及其研究报告撰写的工时工效分析为例进行实验分析,结果表明MASRE系统可有效识别无效工时并进行工时统计与工效分析.同时,该系统已免费向实验室研究团队开放申请注册使用,网址为 .
Keyword ：
任务分配 任务分配 多模态采样 多模态采样 检测方法 检测方法 注意力机制 注意力机制 玩手机行为识别 玩手机行为识别 科研团队 科研团队 科研工效分析 科研工效分析
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | 廖龙龙 , 郑志伟 , 张煜朋 et al. 基于多模态的实验室科研工效分析系统 [J]. | 计算机系统应用 , 2024 , 33 (1) : 68-75 . |
| MLA | 廖龙龙 et al. "基于多模态的实验室科研工效分析系统" . | 计算机系统应用 33 . 1 (2024) : 68-75 . |
| APA | 廖龙龙 , 郑志伟 , 张煜朋 , 方鑫 , 郑育强 , XIONG Ning et al. 基于多模态的实验室科研工效分析系统 . | 计算机系统应用 , 2024 , 33 (1) , 68-75 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
During social interactions, people can obtain a great deal of important information from their tactile senses to improve their relationship with their surroundings. The development of similar capabilities in robots will contribute to the success of intuitive human-robot interaction in the future. In this paper, a tactile sensing method based on the principle of electrical impedance tomography (EIT) is introduced, which with the help of EIT technology and combined with the flexible piezoresistive material Velostat, thin, lightweight, stretchable, and flexible skin can be designed for robots, and at the same time, information about the touch position, duration, and intensity can be acquired, and the image reconstruction is carried out using a dual finite element model, and the experimental results show that based on the flexible The experimental results show that the EIT tactile sensing technology based on the flexible material Velostat can be applied to robotic flexible skin applications. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
Keyword ：
Electric impedance Electric impedance Electric impedance measurement Electric impedance measurement Electric impedance tomography Electric impedance tomography Human robot interaction Human robot interaction Image reconstruction Image reconstruction
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Duan, Zhiqiang , Liu, Lekang , Zhu, Jun et al. A Flexible Tactile Sensor for Robots Based on Electrical Impedance Tomography [C] . 2024 : 123-131 . |
| MLA | Duan, Zhiqiang et al. "A Flexible Tactile Sensor for Robots Based on Electrical Impedance Tomography" . (2024) : 123-131 . |
| APA | Duan, Zhiqiang , Liu, Lekang , Zhu, Jun , Wu, Ruilin , Wang, Yan , Yuan, Xiaohu et al. A Flexible Tactile Sensor for Robots Based on Electrical Impedance Tomography . (2024) : 123-131 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Visual Odometry (VO) has gained significant attention as a critical technology with broad applications in autonomous navigation, augmented reality, and related domains. However, recent research indicates that VO systems are susceptible to adversarial attacks, leading to compromised accuracy and potential system failure. Traditional adversarial attack algorithms, primarily relying on random perturbations or objective function minimization, are not suitable for VO algorithms. In this paper, we present a novel and general adversarial attack algorithm specifically designed for targeting the yaw and translation components of visual odometry, while increasing the Euclidean distance between adjacent frames. Through a comprehensive analysis of VO algorithm characteristics, we propose an effective approach to disrupt VO system operation. Extensive experimental results demonstrate that the proposed attack algorithm significantly reduces the localization accuracy of VO algorithms while exhibiting robustness and generality. The findings of this research contribute to enhancing the security and stability of deep learning-based visual odometry algorithms, providing valuable insights and guidance for practical applications.
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Xie, Xijin , Liao, Longlong , Yu, Yuanlong et al. Universal Adversarial Attacks for Visual Odometry Systems [J]. | 2023 IEEE INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING, ICDL , 2023 : 288-293 . |
| MLA | Xie, Xijin et al. "Universal Adversarial Attacks for Visual Odometry Systems" . | 2023 IEEE INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING, ICDL (2023) : 288-293 . |
| APA | Xie, Xijin , Liao, Longlong , Yu, Yuanlong , Guo, Di , Liu, Huaping . Universal Adversarial Attacks for Visual Odometry Systems . | 2023 IEEE INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING, ICDL , 2023 , 288-293 . |
| Export to | NoteExpress RIS BibTex |
Version ：
Abstract ：
Compressive Sensing for Magnetic Resonance Imaging (CS-MRI) aims to reconstruct Magnetic Resonance (MR) images from under-sampled raw data. There are two challenges to improve CS-MRI methods, i.e. designing an under-sampling algorithm to achieve optimal sampling, as well as designing fast and small deep neural networks to obtain reconstructed MR images with superior quality. To improve the reconstruction quality of MR images, we propose a novel deep convolutional neural network architecture for CS-MRI named MRCSNet. The MRCSNet consists of three sub-networks, a compressive sensing sampling sub-network, an initial reconstruction sub-network, and a refined reconstruction sub-network. Experimental results demonstrate that MRCSNet generates high-quality reconstructed MR images at various under-sampling ratios, and also meets the requirements of real-time CS-MRI applications. Compared to state-of-the-art CS-MRI approaches, MRCSNet offers a significant improvement in reconstruction accuracies, such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity (SSIM). Besides, it reduces the reconstruction error evaluated by the Normalized Root-Mean-Square Error (NRMSE). The source codes are available at https://github.com/TaihuLight/MRCSNet.
Keyword ：
Compressive sensing Compressive sensing magnetic resonance imaging magnetic resonance imaging residual learning residual learning structural similarity index measure structural similarity index measure
Cite：
Copy from the list or Export to your reference management。
| GB/T 7714 | Lu, Hong , Zou, Xiaofei , Liao, Longlong et al. Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images [J]. | INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE , 2021 , 35 (15) . |
| MLA | Lu, Hong et al. "Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images" . | INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE 35 . 15 (2021) . |
| APA | Lu, Hong , Zou, Xiaofei , Liao, Longlong , Li, Kenli , Liu, Jie . Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images . | INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE , 2021 , 35 (15) . |
| Export to | NoteExpress RIS BibTex |
Version ：
Export
| Results： |
Selected to |
| Format： |
