检索结果-内蒙古大学图书馆

A Quality-Aware Voltage Overscaling Framework to Improve the Energy Efficiency and Lifetime of TPUs Based on Statistical Error Modeling

引用

IEEE ACCESS 2024年 12卷 92181-92197页

作者： Senobari, Alireza Vafaei, Jafar Akbari, Omid Hochberger, Christian Shafique, Muhammad Tarbiat Modares Univ Dept Elect & Comp Engn Tehran 14115111 Iran Univ Tehran Sch Elect & Comp Engn Tehran 14395515 Iran Tech Univ Darmstadt Dept Elect Engn D-64289 Darmstadt Germany NYU Abu Dhabi NYU AD Div Engn Abu Dhabi U Arab Emirates

Deep neural networks (DNNs) are a type of artificial intelligence models that are inspired by the structure and function of the human brain, designed to process and learn from large amounts of data, making them particularly well-suited for tasks such as image and speech recognition. However, applications of DNNs are experiencing emerging growth due to the deployment of specialized accelerators such as the Google's Tensor processing Units (TPUs). In large-scale deployments, the energy efficiency of such accelerators may become a critical concern. In the voltage overscaling (VOS) technique, the operating voltage of the system is scaled down beyond the nominal operating voltage, which increases the energy efficiency and lifetime of digital circuits. The VOS technique is usually performed without changing the frequency resulting in timing errors. However, some applications such as multimedia processing, including DNNs, have intrinsic resilience against errors and noise. In this paper, we exploit the inherent resilience of DNNs to propose a quality-aware voltage overscaling framework for TPUs, named X-TPU, which offers higher energy efficiency and lifetime compared to conventional TPUs. The X-TPU framework is composed of two main parts, a modified TPU architecture that supports a runtime voltage overscaling, and a statistical error modeling-based algorithm to determine the voltage of neurons such that the output quality is retained above a given user-defined quality threshold. We synthesized a single-neuron architecture using a 15-nm FinFET technology under various operating voltage levels. Then, we extracted different statistical error models for a neuron corresponding to those voltage levels. Using these models and the proposed algorithm, we determined the appropriate voltage of each neuron (the voltage level of each column of the X-TPU). Results show that running a DNN on X-TPU can achieve 32% energy saving for only 0.6% accuracy loss.

关键词： Voltage control Accuracy artificial neural networks Energy efficiency Approximate computing Power demand Computer architecture Statistical analysis Error analysis Voltage overscaling deep neural networks approximate computing accuracy TPU energy efficiency statistical error analysis

来源：评论

学校读者我要写书评

暂无评论

Tchebichef Transform Domain-Based Deep Learning Architecture for image Super-Resolution

引用

IEEE TRANSACTIONS ON neural networks AND LEARNING SYSTEMS 2024年第2期35卷 2182-2193页

作者： Kumar, Ahlad Singh, Harsh Vardhan Khare, Vijeta Dhirubhai Ambani Inst Informat & Commun Technol D Dept Informat & Commun Technol Gandhinagar 382007 Gujarat India Dhirubhai Ambani Inst Informat & Commun Technol D Gandhinagar 382007 Gujarat India Adani Inst Infrastruct Engn Adalaj 382421 Gujarat India

Recent advances in the area of artificial intelligence and deep learning have motivated researchers to apply this knowledge to solve multipurpose applications in the area of computer vision and image processing. Super-resolution (SR), in the past few years, has produced remarkable results using deep learning methods. The ability of deep learning methods to learn the nonlinear mapping from low-resolution (LR) images to their corresponding high-resolution (HR) images leads to compelling results for SR in diverse areas of research. In this article, we propose a deep learning-based image SR architecture in the Tchebichef transform domain. This is achieved by integrating a transform layer into the proposed architecture through a customized Tchebichef convolutional layer (TCL). The role of TCL is to convert the LR image from the spatial domain to the orthogonal transform domain using Tchebichef basis functions. The inversion of the transform mentioned earlier is achieved using another layer known as the inverse TCL (ITCL), which converts back the LR images from the transform domain to the spatial domain. It has been observed that using the Tchebichef transform domain for the task of SR takes the advantage of high and low-frequency representation of images that makes the task of SR simplified. Furthermore, a transfer learning-based approach is adopted to enhance the quality of images by considering Covid19 medical images as an additional experiment. It is shown that our architecture enhances the quality of X-ray and CT images of COVID-19, providing a better image quality that may help in clinical diagnosis. Experimental results obtained using the proposed Tchebichef transform domain SR (TTDSR) architecture provides competitive results when compared with most of the deep learning methods employed using a fewer number of trainable parameters.

关键词： Transforms Deep learning Computer architecture Convolutional neural networks COVID-19 Training Kernel Convolutional neural network (CNN) COVID-19 deep learning image super-resolution (SR) Tchebichef moments

来源：评论

学校读者我要写书评

暂无评论

Explainable AI for Medical Data: Current Methods, Limitations, and Future Directions

引用

ACM COMPUTING SURVEYS 2025年第6期57卷 1-46页

作者： Hossain, Md Imran Zamzmi, Ghada Mouton, Peter r. Salekin, Md Sirajus Sun, Yu Goldgof, Dmitry Univ S Florida Dept Comp Sci & Engn ENB 4202 E Fowler Ave Tampa FL 33620 USA SRC Biosci 1810 W Kennedy Blvd Tampa FL 33606 USA

With the power of parallel processing, large datasets, and fast computational resources, deep neural networks (DNNs) have outperformed highly trained and experienced human experts in medical applications. However, the large global community of healthcare professionals, many of whom routinely face potentially life-or-death outcomes with complex medicolegal consequences, have yet to embrace this powerful technology. The major problem is that most current AI solutions function as a metaphorical black-box positioned between input data and output decisions without a rigorous explanation for their internal processes. With the goal of enhancing trust and improving acceptance of artificial intelligence- (AI) based technology in clinical medicine, there is a large and growing effort to address this challenge using eXplainable AI (XAI), a set of techniques, strategies, and algorithms with an explicit focus on explaining the "hows and whys" of DNNs. Here, we provide a comprehensive review of the state-of-the-art XAI techniques concerning healthcare applications and discuss current challenges and future directions. We emphasize the strengths and limitations of each category, including image, tabular, and textual explanations, and explore a range of evaluation metrics for assessing the effectiveness of XAI solutions. Finally, we highlight promising opportunities for XAI research to enhance the acceptance of DNNs by the healthcare community.

关键词： Explainability medical data responsible AI deep neural networks pretable AI

来源：评论

学校读者我要写书评

暂无评论

Deep spiking neural networks based on model fusion technology for remote sensing image classification

引用

ENGINEERING applications OF artificial INTELLIGENCE 2025年 142卷

作者： Niu, Li-Ye Wei, Ying Zhao, Liping Hu, Keli Zhejiang Chinese Med Univ Zhejiang Prov Hosp Chinese Med Dept Radiol Affiliated Hosp 1 Hangzhou 310006 Peoples R China Northeastern Univ Coll Informat Sci & Engn Shenyang 110819 Peoples R China Shaoxing Univ Dept Comp Sci & Engn Shaoxing 312000 Peoples R China

The spiking neural network (SNN) based on brain inspiration, as the third-generation neural network, has attracted great research interest due to its ultra-low power event-driven data processing method. How to obtain high-accuracy deep networks has always been a challenge in the field of SNN. At present, there are two main methods for training deep spiking neural networks (SNNs). The first is the spike-based spatiotemporal backpropagation (STBP), and the other indirect is to convert the trained artificial neural network (ANN) into SNN (ANN-SNN). Algorithms that directly train SNNs are usually inefficient, while the ANN-SNN-based models require along inference time and also suffer from performance losses. The model fusion technology (MFT) proposed in this paper, combined with ANN-SNN and STBP, provides a new training paradigm for obtaining deep SNNs. We propose a bilateral multi-strength integrate-and-fire (BM-IF) spiking neuron for ANN-SNN to simplify the conversion operation. Under the same network architecture and encoding time window, the ResNet34 based on our algorithm achieved State-of-the-art (SOTA) in imageNet. At the same time, we combined transfer learning methods to achieve SOTA results on three remote sensing scene classification datasets. Our results also indicate that the MFT-based SNN outperforms the optimal inference accuracy provided by ANNSNN in terms of network architecture and dataset inference time steps reduced by 512-1250 times. It can be seen that this algorithm provides a low-latency and high-precision training scheme for the development of deep SNN.

关键词： Spiking neural network Model fusion technology Low-latency High-precision Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Proton-Gated Synaptic Transistors, Based on an Electron-Beam Patterned Nafion Electrolyte

引用

ACS APPLIED MATERIALS & INTERFACES 2023年第15期15卷 19279-19289页

作者： Mohanty, Himadri Nandan Tsuruoka, Tohru Mohanty, Jyoti Ranjan Terabe, Kazuya Indian Inst Technol Hyderabad Dept Phys Nanomagnetism & Microscopy Lab Sangareddy 502285 Telangana India Natl Inst Mat Sci Res Ctr Mat Nanoarchitecton Tsukuba Ibaraki 305004 Japan

Neuromorphic processors using artificial neural networks are the center of attention for energy-efficient analog computing. artificial synapses act as building blocks in such neural networks for parallel information processing and data storage. Herein we describe the fabrication of a proton-gated synaptic transistor using a Nafion electrolyte thin film, which is patterned by electron-beam lithography (EBL). The device has an active channel of indium???zinc-oxide (IZO) between the source and drain electrodes, which shows Ohmic behavior with a conductance level on the order of 100 ??S. Under voltage applications to the gate electrode, the channel conductance is changed due to the injection and extraction of protons between the IZO channel and the Nafion electrolyte, emulating various synaptic functions with short-term and long-term plasticity. When positive (negative) gate voltage pulses are consecutively applied, the device exhibits long-term potentiation (depression) at the same number of steps as the number of input pulses. Based on these characteristics, an artificial neural network using this transistor shows ???84% image recognition accuracy for handwritten digits. The subject transistor also successfully mimics paired-pulse facilitation and depression, Hebbian spike-timing-dependent plasticity, and Pavlovian associative learning followed by extinction activities. Finally, dynamical pattern image memorization is demonstrated in a 5 ?? 5 array of these synaptic transistors. The results indicate that EBL patternable Nafion electrolytes have great potential for use in the fabrication and circuit-level integration of synaptic devices for neuromorphic computing applications.

关键词： protonic transistor Nafion electron-beam patterning synaptic plasticity Hebbian and Pavlovian learning artificial neural network

来源：评论

学校读者我要写书评

暂无评论

Patch- and Class-Wise Hyperspectral Knowledge Learning: A Composite Consistency-Constrained Self-Ensemble Framework for Change Detection

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2025年 63卷

作者： Zhao, Xiaoyang Li, Siyao Liu, Xinyue Song, Chuanming Wang, Xianghai Liaoning Normal Univ Sch Geog Sci Dalian 116029 Peoples R China Liaoning Normal Univ Sch Comp & Artificial Intelligence Dalian 116029 Peoples R China Dalian Univ Sch Informat Engn Dalian 116622 Peoples R China

Obtaining fine land surface change information from multitemporal hyperspectral images (HSIs) is a key goal pursued in remote sensing image processing. Recently, HSI change detection (HSI-CD) methods based on convolutional neural networks (CNNs) have achieved surprising detection results. One of the reasons is the support of large-scale labeled samples for network learning. However, the existence of mixed pixels greatly increases the difficulty of HSI interpretation, resulting in accurate pixel-level labeling work with a heavy burden and unable to meet the needs of time-sensitive applications. For this reason, achieving stable and high-precision CD with fewer samples is a difficult issue in this field. To address the above problems, a composite consistency-constrained self-ensemble framework (C3SelF) for HSI-CD is proposed, to alleviate the problems of low detection accuracy and instability caused by small samples. The framework mainly comprises two lightweight networks with the same structure aiming at accelerating the model inference process and thus improving the processing timeliness. The composite learning mode implements patch-wise classification loss, class-wise consistency loss on labeled samples, and patch-wise consistency loss on unlabeled samples under a multilevel noise perturbation strategy, which improves the classification results and reduces the labeling cost. Moreover, to exploit the multidimensional features contained in HSIs, a lightweight selective spatial-spectral feature joint network (S3Net) is designed to overcome over-fitting, and to deeply mine the discriminative information in unlabeled samples, a new sample screening strategy is designed to ensure the stability of the network during training unlabeled samples. Extensive experiments prove that the proposed C3SelF outperforms the state-of-the-art (SOTA) methods at a sampling rate of 0.1%, reaching 93.44% Kappa and 97.26% overall accuracy (OA) on the Farmland dataset. The source code of the

关键词： Accuracy Training Feature extraction Hyperspectral imaging Transformers Data mining Vectors Semisupervised learning Convolutional neural networks Unsupervised learning Change detection (CD) convolutional neural networks (CNNs) deep learning hyperspectral image (HSI) processing semi-supervised learning

来源：评论

学校读者我要写书评

暂无评论

TinySpiking: a lightweight and efficient python framework for unsupervised learning spiking neural networks

ENGINEERING RESEARCH EXPRESS

引用

ENGINEERING RESEARCH EXPRESS 2025年第1期7卷

作者： Liu, Xin Mo, Lingfei Tang, Mengting Southeast Univ Sch Instrument Sci & Engn FutureX LAB Nanjing Peoples R China

neural computation frameworks are essential for advancing computational neuroscience and artificial intelligence, offering a robust platform for simulating intricate brain-like processes and fostering the growth of intelligent systems. This study presents TinySpiking, a novel, lightweight, and energy-efficient Python framework designed for the simulation and learning of Spiking neural networks (SNNs). Unlike traditional frameworks, TinySpiking does not depend on third-party libraries, thereby reducing computational overhead and energy consumption. The framework's innovation is that it implements unsupervised learning through Spike-Time Dependent Plasticity (STDP), a biologically inspired learning rule that adjusts synaptic weights based on the precise timing of neuronal spikes. This mechanism is crucial for constructing complex neural architectures and effectively processing spatio-temporal data, which is vital for the dynamic and intricate demands of real-world data analysis. Demonstrating the practical utility of TinySpiking, we applied it to image reconstruction tasks using the MNIST and landmark datasets. The results not only validate the network's ability to autonomously identify and enhance key visual features without external supervision but also highlight its efficiency in learning from data, mirroring the adaptability of biological neural systems. In conclusion, TinySpiking's innovative, lightweight design and its proven effectiveness in unsupervised learning tasks make it a standout computational tool for the fields of computational neuroscience and machine learning. Its low time consumption, biological plausibility, and independence from third-party libraries position it as a compelling platform for future research and applications, promising to drive advancements in unsupervised learning and intelligent system development.

关键词： spiking neural networks STDP unsupervised learning image reconstruction computational neuroscience

来源：评论

学校读者我要写书评

暂无评论

Solution-processed self-assembling charge-transfer cocrystal/TIPS-pentacene heterojunctions for artificial synapses

引用

JOURNAL OF MATERIALS CHEMISTRY C 2025年第22期13卷 11357-11365页

作者： Jin, Tao Li, Wenju Zhang, Yongyi Wang, Ruiheng Wang, Shuai Pan, Chen Wang, Guan Zhang, Jiacheng Yao, Lei Zhang, Jing Zhang, Qichun Nanjing Univ Posts & Telecommun State Key Lab Flexible Elect LoFE 9 Wenyuan Rd Nanjing 210023 Peoples R China Nanjing Univ Posts & Telecommun Inst Adv Mat IAM 9 Wenyuan Rd Nanjing 210023 Peoples R China City Univ Hong Kong Dept Mat Sci & Engn Hong Kong Peoples R China

artificial synapses have gained considerable attention in modeling artificial visual systems. A facile and efficient p-n heterojunction, acting as the essential component, plays a critical role in configuration optimization and processing capabilities. Herein, we introduced a novel charge-transfer cocrystal that consisted of 5,8-dihydroindolo[2,3-c]carbazole (DDCB) and 7,7,8,8-tetracyanoquinodimethane (TCNQ) via the solution-process, which demonstrated good crystallinity and typical n-type charge transport characteristics. To realize good charge storage ability, crystalline p-n heterojunctions with a phase-separated heterostructure were one-step prepared by drop-casting TIPS-pentacene and DDCB/TCNQ solutions onto the substrate. The as-prepared artificial synaptic devices exhibited remarkable paired-pulse facilitation (PPF), spike-duration dependent plasticity, spike-intensity dependent plasticity, and learning-experience behaviors similar to those of biological synapses in response to light pulses. Furthermore, based on the light potentiation and electrical depression modulation properties of these synaptic devices, we constructed artificial neural networks for image recognition, achieving a validation accuracy of over 97% on the modified national institute of standards and technology (MNIST) dataset. This work provides a new simple approach to establishing p-n heterojunction based artificial synapses for applications in light information processing, artificial vision, and memory simulation.

关键词： image coding

来源：评论

学校读者我要写书评

暂无评论

DLSIA: Deep Learning for Scientific image Analysis

引用

JOURNAL OF APPLIED CRYSTALLOGRAPHY 2024年第2期57卷 392-402页

作者： Roberts, Eric J. Chavez, Tanny Hexemer, Alexander Zwart, Petrus H. Lawrence Berkeley Natl Lab Ctr Adv Math Energy Res Applicat Berkeley CA 94720 USA Lawrence Berkeley Natl Lab Mol Biophys & Integrated Bioimaging Div Berkeley CA 94720 USA Lawrence Berkeley Natl Lab Adv Light Source Berkeley CA 94720 USA Lawrence Berkeley Natl Lab Berkeley Synchrotron Infrared Struct Biol Program Berkeley CA 94720 USA

DLSIA (Deep Learning for Scientific image Analysis) is a Python-based machine learning library that empowers scientists and researchers across diverse scientific domains with a range of customizable convolutional neural network (CNN) architectures for a wide variety of tasks in image analysis to be used in downstream data processing. DLSIA features easy-to-use architectures, such as autoencoders, tunable U-Nets and parameter-lean mixed-scale dense networks (MSDNets). Additionally, this article introduces sparse mixed-scale networks (SMSNets), generated using random graphs, sparse connections and dilated convolutions connecting different length scales. For verification, several DLSIA-instantiated networks and training scripts are employed in multiple applications, including inpainting for X-ray scattering data using U-Nets and MSDNets, segmenting 3D fibers in X-ray tomographic reconstructions of concrete using an ensemble of SMSNets, and leveraging autoencoder latent spaces for data compression and clustering. As experimental data continue to grow in scale and complexity, DLSIA provides accessible CNN construction and abstracts CNN complexities, allowing scientists to tailor their machine learning approaches, accelerate discoveries, foster interdisciplinary collaboration and advance research in scientific image analysis.

关键词： deep learning convolutional neural networks X-ray scattering tomography data compression

来源：评论

学校读者我要写书评

暂无评论

AutoFace: How to Obtain Mobile neural Network-Based Facial Feature Extractor in Less Than 10 Minutes?

引用

IEEE ACCESS 2024年 12卷 25106-25118页

作者： Savchenko, Andrey V. Sber AI Lab Moscow 117312 Russia HSE Univ Lab Algorithms & Technol Network Anal Nizhnii Novgorod 603155 Russia

Various mobile and edge devices have significantly different processing capabilities, making it challenging to develop a single universal architecture of a neural network to extract facial embeddings. In this paper, we study the automated machine learning techniques to design a neural network with the best performance on a concrete device. The novel procedure is proposed to choose the better subnetwork of the Supernet based on a genetic algorithm with a surrogate binary classifier to compare the expected accuracy of two subnetworks. The latter uses only encoding of a candidate subnetwork and does not require directly estimating its accuracy on a validation set. As a result, the most computationally efficient and accurate model in TensorFlow Lite format is obtained in less than 10 minutes for a specific device and latency constraint. An Android demo application has been developed to demonstrate the potential of designed neural networks. It is experimentally shown that the proposed approach is universal: it can extract deep embeddings for tasks such as face verification and facial expression recognition and for various types of devices, including smartphones and Raspberry Pi single-board mini-computers. Our models process one facial image in real-time and achieve much higher accuracy when compared to the best-known lightweight networks.

关键词： Face recognition Task analysis Training Computer architecture Feature extraction artificial neural networks Performance evaluation Mobile communication Genetic algorithms Search methods Evolutionary computation Emotion recognition Mobile applications edge devices genetic algorithms evolutionary search face verification facial expression recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：