检索结果-南通市图书馆

Geometric prior guided hybrid deep neural network for facial beauty analysis

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

CAAI Transactions on Intelligence Technology 2024年第2期9卷 467-480页

作者： Tianhao Peng Mu Li Fangmei Chen Yong Xu david zhang The School of Computer Science and Technology Guizhou UniversityGuiyangChina Department of Automation Moutai InstituteRenhuaiGuizhouChina The School of Computer Science and Technology Harbin Institute of TechnologyShenzhenShenzhenChina The Information and Communication Engineering Department Dalian Minzu UniversityDalianChina The School of Data Science The Chinese University of Hong KongShenzhenShenzhenChina

Facial beauty analysis is an important topic in human *** may be used as a guidance for face beautification applications such as cosmetic *** neural networks(DNNs)have recently been adopted for facial beauty analysis and have achieved remarkable ***,most existing DNN-based models regard facial beauty analysis as a normal classification *** ignore important prior knowledge in traditional machine learning models which illustrate the significant contribution of the geometric features in facial beauty *** be specific,landmarks of the whole face and facial organs are introduced to extract geometric features to make the *** by this,we introduce a novel dual-branch network for facial beauty analysis:one branch takes the Swin Transformer as the backbone to model the full face and global patterns,and another branch focuses on the masked facial organs with the residual network to model the local patterns of certain facial ***,the designed multi-scale feature fusion module can further facilitate our network to learn complementary semantic information between the two *** model optimisation,we propose a hybrid loss function,where especially geometric regulation is introduced by regressing the facial landmarks and it can force the extracted features to convey facial geometric *** performed on the SCUT-FBP5500 dataset and the SCUT-FBP dataset demonstrate that our model outperforms the state-of-the-art convolutional neural networks models,which proves the effectiveness of the proposed geometric regularisation and dual-branch structure with the hybrid *** the best of our knowledge,this is the first study to introduce a Vision Transformer into the facial beauty analysis task.

关键词： deep neural networks face analysis face biometrics image analysis

Organic heterojunction synaptic device with ultra high recognition rate for neuromorphic computing

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Nano Research 2024年第6期17卷 5614-5620页

作者： Xuemeng Hu Jialin Meng Tianyang Feng Tianyu Wang Hao Zhu Qingqing Sun david Wei zhang Lin Chen School of Microelectronics State Key Laboratory of Integrated Chips and SystemsFudan UniversityShanghai 200433China National Integrated Circuit Innovation Center Shanghai 201203China

Traditional computing structures are blocked by the von Neumann bottleneck,and neuromorphic computing devices inspired by the human brain which integrate storage and computation have received more and more ***,a flexible organic device with 2,7-dioctyl[1]benzothieno[3,2-b][1]benzothiophene(C8-BTBT)and 2,9-didecyldinaphtho[2,3-b:2′,3′-f]thieno[3,2-b]thiophene(C10-DNTT)heterostructural channel having excellent synaptic behaviors was fabricated on muscovite(MICA)substrate,which has a memory window greater than 20 *** device shows better electrical characteristics than organic field effect transistors with single organic semiconductor ***,the device simulates organism synaptic behaviors successfully,such as paired-pulse facilitation(PPF),long-term potentiation/depression(LTP/LTD)process,and transition from short-term memory(STM)to long-term memory(LTM)by optical and electrical ***,the neuromorphic computing function was verified using the Modified National Institute of Standards and Technology(MNIST)pattern recognition,with a recognition rate nearly 100%without *** research proposes a flexible organic heterojunction with the ultra-high recognition rate in MNIST pattern recognition and provides the possibility for future flexible wearable neuromorphic computing devices.

关键词： organic heterojunction neuromorphic computing synapse behaviors optical modulation Modified National Institute of Standards and Technology(MNIST)pattern recognition

InGaZnO-based photoelectric synaptic devices for neuromorphic computing

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Semiconductors 2024年第9期45卷 42-47页

作者： Jieru Song Jialin Meng Tianyu Wang Changjin Wan Hao Zhu Qingqing Sun david Wei zhang Lin Chen School of Microelectronics State Key Laboratory of Integrated Chips and SystemsFudan UniversityShanghai 200433China School of Integrated Circuits Shandong UniversityJinan 250100China National Integrated Circuit Innovation Center Shanghai 201203China School of Electronic Science and Engineering Nanjing UniversityNanjing 210023China Jiashan Fudan Institute Jiashan 314102China

Photoelectric synaptic devices could emulate synaptic behaviors utilizing photoelectric effects and offer promising prospects with their high-speed operation and low crosstalk. In this study, we introduced a novel InGaZnO-based photoelectric memristor. Under both electrical and optical stimulation, the device successfully emulated synaptic characteristics including excitatory postsynaptic current (EPSC), paired-pulse facilitation (PPF), long-term potentiation (LTP), and long-term depression (LTD). Furthermore, we demonstrated the practical application of our synaptic devices through the recognition of handwritten digits. The devices have successfully shown their ability to modulate synaptic weights effectively through light pulse stimulation, resulting in a recognition accuracy of up to 93.4%. The results illustrated the potential of IGZO-based memristors in neuromorphic computing, particularly their ability to simulate synaptic functionalities and contribute to image recognition tasks.

关键词： InGaZnO artificial synapse neuromorphic computing photoelectric memristor

Population-influenced spatiotemporal pattern of natural disaster and social crisis in China, AD1–1910

在线全文

学校读者我要写书评

暂无评论

Science China Earth Sciences 2019年第7期62卷 1138-1150页

作者： Shengda zhang david Dian zhang Department of Geography The University of Hong Kong School of Geographical Sciences Guangzhou University

Researchers mostly ascribe contemporary natural disaster and the concomitant social crisis to anomalous climate change or global warming. However, whether such a relationship is still valid in long-term historical settings remains doubtful. In this study, data obtained from historical records about natural calamities(flood and drought) and their social impacts(famine,cannibalism, and war) at provincial and decadal scales during AD1–1910 are applied to mathematical statistics such as correlation and regression analysis as well as spatial visualization. Also, the role of population in the nexus of meteorological catastrophes and human miseries is investigated. Results show that at the provincial scale, generally there is high consistency among different variables and most of them are clustered in eastern part of China, especially in the north. More in-depth examinations indicate regional disparities that variables account for higher proportion in the south during the later imperial era,which may be attributed to the southward movement of population center. At the decadal scale, drought is the primary contributor to famine and cannibalism, while severe or even out-of-control famine i.e. cannibalism is more likely to incur war than ordinary famine per se. Besides, population growth rate exerts its positive effect on natural hazards and food crises, whereas more wars bring population loss and relieve population stress despite the weak negative association. These findings supplement previous views and confirm that the distributions of disasters in ancient China are affected by population rather than climatic variability, which only determines the occurrence of disaster.

关键词： Natural disaster Social crisis Population Spatiotemporal pattern Historical China

Fast and convergence-guaranteed algorithm for linear separation

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2010年第4期53卷 729-737页

作者： zhang david Center for Multimedia Signal Processing/Department of Computing Hong Kong Polytechnic University

Efficient linear separation algorithms are important for pattern classification applications. In this paper, an algorithm is developed to solve linear separation problems in n-dimensional space. Its convergence feature is proved. The proposed algorithm is proved to converge to a correct solution whenever the two sets are separable. The complexity of the proposed algorithm is analyzed, and experiments on both randomly generated examples and real application problems were carried out. While analysis shows that its time complexity is lower than SVM that needs computations for quadratic programming optimization, experiment results show that the developed algorithm is more efficient than the least-mean-square (LMS), and the Perceptron.

关键词： linear separation problem classification convergence efficiency complexity

Two-dimensional materials for synaptic electronics and neuromorphic systems

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science Bulletin 2019年第15期64卷 1056-1066页

作者： Shuiyuan Wang david Wei zhang Peng Zhou State Key Laboratory of ASIC and System School of Microelectronics Fudan University

Synapses in biology provide a variety of functions for the neural system. Artificial synaptic electronics that mimic the biological neuron functions are basic building blocks and developing novel artificial synapses is essential for neuromorphic computation. Inspired by the unique features of biological synapses that the basic connection components of the nervous system and the parallelism, low power consumption, fault tolerance, self-learning and robustness of biological neural systems, artificial synaptic electronics and neuromorphic systems have the potential to overcome the traditional von Neumann bottleneck and create a new paradigm for dealing with complex problems such as pattern recognition, image classification, decision making and associative learning. Nowadays, two-dimensional(2 D) materials have drawn great attention in simulating synaptic dynamic plasticity and neuromorphic computing with their unique properties. Here we describe the basic concepts of bio-synaptic plasticity and learning, the 2 D materials library and its preparation. We review recent advances in synaptic electronics and artificial neuromorphic systems based on 2 D materials and provide our perspective in utilizing 2 D materials to implement synaptic electronics and neuromorphic systems in hardware.

关键词： Artificial synaptic electronics Neuromorphic computation 2D materials Synaptic plasticity Hebbian learning

达格列净在血糖控制不佳肥胖2型糖尿病患者临床治疗中的效果观察

在线全文

学校读者我要写书评

暂无评论

反射疗法与康复医学 2024年第16期5卷 75-77+92页

作者：岳晓燕张大卫荣成市人民医院内分泌风湿病科荣成市人民医院心内科

目的探讨达格列净在血糖控制不佳肥胖2型糖尿病患者临床治疗中的效果。方法选择2022年10月—2023年10月该院收治的104例肥胖2型糖尿病患者为研究对象，按随机数字表法将其分为对照组与观察组，各52例。对照组采用二甲双胍缓释片治疗... 详细信息

目的探讨达格列净在血糖控制不佳肥胖2型糖尿病患者临床治疗中的效果。方法选择2022年10月—2023年10月该院收治的104例肥胖2型糖尿病患者为研究对象，按随机数字表法将其分为对照组与观察组，各52例。对照组采用二甲双胍缓释片治疗，观察组采用二甲双胍缓释片联合达格列净治疗。比较两组血糖指标、血脂指标、胰岛功能指标、身体质量指数、腹围指标及不良反应发生情况。结果治疗前，两组血糖指标、血脂指标、胰岛功能指标、身体质量指数及腹围指标比较，组间差异无统计学意义（P>0.05）：治疗后，观察组空腹血糖、糖化血红蛋白、总胆固醇、甘油三酯、低密度脂蛋白胆固醇、胰岛素抵抗指数、身体质量指数及腰围指标水平均低于对照组，胰岛β细胞功能指数、空腹胰岛素、空腹C肽水平均高于对照组，组间差异有统计学意义（P0.05）。结论达格列净治疗效果优于利格列汀，其降血糖、血脂及改善胰岛功能、减轻患者体重及缩小腹围的作用更佳，且安全性良好。

关键词： 2型糖尿病肥胖二甲双胍缓释片达格列净

A non-volatile AND gate based on Al2O3/HfO2/Al2O3 charge-trap stack for in-situ storage applications

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science Bulletin 2019年第20期64卷 1518-1524页

作者： Jingyu Li Heng zhang Yi Ding Jiayi Li Shuiyuan Wang david Wei zhang Peng Zhou State Key Laboratory of ASIC and System School of Microelectronics Fudan University

The emergence of two-dimensional(2 D) materials has inspired academia in microelectronics to find a novel device structure to offer a potential technological route for increasing energy efficiency and processing speed for data storage and computing at transistor level. Devices based on 2 D materials include logic gates and memories, each with their own unique features. However, integrating logic function and memory into a single device has barely been studied. Here, we report a non-volatile AND gate based on Al2O3/HfO2/Al2O3 charge-trap stack. The device can store charges after completing logic operation. The ratio of high and low current states during logic operations can exceed 105. The output current states during the logic and memory operations still have a two orders of magnitude distinction after 800 s, indicating that this device possesses the non-volatile characteristic. The device has potential applications for insitu memory applications, which makes it a possible candidate to break the ‘‘memory wall' at transistor level.

关键词： AND gate Non-volatile memory Charge-trap gate stack In-situ memory

New evidence for human occupation of the northern Tibetan Plateau,China during the Late Pleistocene

在线全文

学校读者我要写书评

暂无评论

Chinese Science Bulletin 2007年第19期52卷 2675-2679页

作者： YUAN BaoYin HUANG WeiWen zhang david Institute of Geology and Geophysics Chinese Academy of Sciences Beijing 100029 China Institute of Vertebrate Paleontology and Paleoanthropology Chinese Academy of Sciences Beijing 100044 China Department of Geography Hong Kong University Hong Kong China

The described stone artifacts are recovered from the 70 m-high terrace (4600 m a.s.l.) at the southeastern shore of the Siling Co on the northern Tibetan Plateau. The terrace was formed during the Interstadial period before the LGM, ca. 40-30 ka B.P. based on paleoenvironmental research. The Paleoliths from the Siling Co provide evidence for early human occupation of the northern Tibetan Plateau. They show technological and typological affinities with the European Middle Paleolithic suggesting that the early human occupation here might relate to migratory waves during the Late Pleistocene that dispersed humans across the Old World.

关键词：青藏高原微波旧石器时代地质