检索结果-南通市图书馆

FSCIL-EACA: Few-Shot Class-Incremental Learning network Based on Embedding Augmentation and Classifier Adaptation for Image Classification

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第1期33卷 139-152页

作者： Ruru ZHANG Haihong E Meina SONG School of Computer Science Beijing University of Posts and Telecommunications Education department information network engineering Research Center Beijing University of Posts and Telecommunications

The ability to learn incrementally is critical to the long-term operation of AI systems. Benefiting from the power of few-shot class-incremental learning(FSCIL), deep learning models can continuously recognize new classes with only a few samples. The difficulty is that limited instances of new classes will lead to overfitting and exacerbate the catastrophic forgetting of the old classes. Most previous works alleviate the above problems by imposing strong constraints on the model structure or parameters, but ignoring embedding network transferability and classifier adaptation(CA), failing to guarantee the efficient utilization of visual features and establishing relationships between old and new classes. In this paper, we propose a simple and novel approach from two perspectives: embedding bias and classifier bias. The method learns an embedding augmented(EA) network with cross-class transfer and class-specific discriminative abilities based on self-supervised learning and modulated attention to alleviate embedding bias. Based on the adaptive incremental classifier learning scheme to realize incremental learning capability,guiding the adaptive update of prototypes and feature embeddings to alleviate classifier bias. We conduct extensive experiments on two popular natural image datasets and two medical datasets. The experiments show that our method is significantly better than the baseline and achieves state-of-the-art results.

关键词： Training Adaptation models Visualization Image recognition Prototypes Self-supervised learning Power capacitors

Fairness in machine learning: definition, testing,debugging, and application

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 China(information Sciences) 2024年第09期67卷 41-61页

作者： Xuanqi GAO Chao SHEN Weipeng JIANG Chenhao LIN Qian LI Qian WANG Qi LI Xiaohong GUAN Faculty of Electronic and information engineering Xi'an Jiaotong University School of Cyber Science and engineering Wuhan University Institute for network Sciences and Cyberspace Tsinghua University

In recent years, artificial intelligence technology has been widely used in many fields, such as computer vision, natural language processing and autonomous driving. Machine learning algorithms, as the core technique of AI, have significantly facilitated people's lives. However, underlying fairness issues in machine learning systems can pose risks to individual fairness and social security. Studying fairness definitions, sources of problems, and testing and debugging methods of fairness can help ensure the fairness of machine learning systems and promote the wide application of artificial intelligence technology in various fields. This paper introduces relevant definitions of machine learning fairness and analyzes the sources of fairness problems. Besides, it provides guidance on fairness testing and debugging methods and summarizes popular datasets. This paper also discusses the technical advancements in machine learning fairness and highlights future challenges in this area.

关键词： artificial intelligence security machine learning security machine learning fairness model testing model debugging

同方期刊数据库

MarkNeRF:Watermarking for Neural Radiance Field

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第7期80卷 1235-1250页

作者： Lifeng Chen Jia Liu Wenquan Sun Weina Dong Xiaozhong Pan Cryptographic engineering department Institute of Cryptographic EngineeringEngineering University of PAPXi’an710086China Key Laboratory of network and information Security of PAP Xi’an710086China

This paper presents a novel watermarking scheme designed to address the copyright protection challenges encountered with Neural radiation field(NeRF)*** employ an embedding network to integrate the watermark into the images within the training ***,theNeRFmodel is utilized for *** copyright verification,a secret image is generated by inputting a confidential viewpoint into *** this basis,design an extraction network to extract embedded watermark images fromconfidential *** the event of suspicion regarding the unauthorized usage of NeRF in a black-box scenario,the verifier can extract the watermark from the confidential viewpoint to authenticate the model’s *** experimental results demonstrate not only the production of visually appealing watermarks but also robust resistance against various types of noise attacks,thereby substantiating the effectiveness of our approach in safeguarding NeRF.

关键词： Neural radiation field 3D watermark robustness black box

Multi-instance partial-label learning: towards exploiting dual inexact supervision

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

Science China(information Sciences) 2024年第3期67卷 48-61页

作者： Wei TANG Weijia ZHANG Min-Ling ZHANG School of Computer Science and engineering Southeast University Key Laboratory of Computer network and information Integration (Southeast University) Ministry of Education School of information and Physical Sciences The University of Newcastle

Weakly supervised machine learning algorithms are able to learn from ambiguous samples or labels, e.g., multi-instance learning or partial-label learning. However, in some real-world tasks, each training sample is associated with not only multiple instances but also a candidate label set that contains one groundtruth label and some false positive labels. Specifically, at least one instance pertains to the ground-truth label while no instance belongs to the false positive labels. In this paper, we formalize such problems as multi-instance partial-label learning(MIPL). Existing multi-instance learning algorithms and partial-label learning algorithms are suboptimal for solving MIPL problems since the former fails to disambiguate a candidate label set, and the latter cannot handle a multi-instance bag. To address these issues, a tailored algorithm named MIPLGP, i.e., multi-instance partial-label learning with Gaussian processes, is *** first assigns each instance with a candidate label set in an augmented label space, then transforms the candidate label set into a logarithmic space to yield the disambiguated and continuous labels via an exclusive disambiguation strategy, and last induces a model based on the Gaussian processes. Experimental results on various datasets validate that MIPLGP is superior to well-established multi-instance learning and partial-label learning algorithms for solving MIPL problems.

关键词： machine learning multi-instance partial-label learning multi-instance learning partial-label learning Gaussian processes

Gradient sparsification for efficient wireless federated learning with differential privacy

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 China(information Sciences) 2024年第4期67卷 272-288页

作者： Kang WEI Jun LI Chuan MA Ming DING Feng SHU Haitao ZHAO Wen CHEN Hongbo ZHU School of Electronic and Optical engineering Nanjing University of Science and Technology Zhejiang Lab Key Laboratory of Computer network and information Integration(Southeast University) Ministry of Education Data61 Commonwealth Scientific and Industrial Research Organisation School of information and Communication engineering Hainan University School of Communications and information engineering Nanjing University of Posts and Telecommunications School of Electronic information and Electrical engineering Shanghai Jiao Tong University

Federated learning(FL) enables distributed clients to collaboratively train a machine learning model without sharing raw data with each other. However, it suffers from the leakage of private information from uploading models. In addition, as the model size grows, the training latency increases due to the limited transmission bandwidth and model performance degradation while using differential privacy(DP)protection. In this paper, we propose a gradient sparsification empowered FL framework with DP over wireless channels, to improve training efficiency without sacrificing convergence performance. Specifically, we first design a random sparsification algorithm to retain a fraction of the gradient elements in each client’s local model, thereby mitigating the performance degradation induced by DP and reducing the number of transmission parameters over wireless channels. Then, we analyze the convergence bound of the proposed algorithm, by modeling a non-convex FL problem. Next, we formulate a time-sequential stochastic optimization problem for minimizing the developed convergence bound, under the constraints of transmit power, the average transmitting delay, as well as the client’s DP requirement. Utilizing the Lyapunov drift-plus-penalty framework, we develop an analytical solution to the optimization problem. Extensive experiments have been implemented on three real-life datasets to demonstrate the effectiveness of our proposed algorithm. We show that our proposed algorithms can fully exploit the interworking between communication and computation to outperform the baselines, i.e., random scheduling, round robin, and delay-minimization algorithms.

关键词： federated learning differential privacy gradient sparsification Lyapunov drift convergence analysis

engineering the spectra of photon triplets generated from micro/nanofiber

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Physics B 2024年第3期33卷 408-414页

作者：瞿川郭东琴李笑笑刘振旗赵义张胜海卫正统 The College of Basic department Information Engineering UniversityZhengzhou 450000China

Quantum light sources are the core resources for photonics-based quantum information *** investigate the spectral engineering of photon triplets generated by third-order spontaneous parametric down-conversion in micro/*** phase mismatching at one-third pump frequency gives rise to non-degenerate photon triplets,the joint spectral intensity of which has an elliptical locus with a fixed eccentricity of√6/***,we propose a frequency-division scheme to separate non-degenerate photon triplets into three channels with high heralding efficiency for the first *** an appropriate pump wavelength can compensate for the fabrication errors of micro/nanofiber and also generate narrowband,non-degenerate photon triplet sources with a high signal-to-noise ***,the long-period micro/nanofiber grating introduces a new controllable degree of freedom to tailor phase matching,resulting from the periodic oscillation of *** this scheme,the wavelength of photon triplets can be flexibly tuned using quasi-phase *** study the generation of photon triplets from this novel perspective of spectrum engineering,and we believe that this work will accelerate the practical implementation of photon triplets in quantum information processing.

关键词： photon triplets micro/nanofiber spectrum engineering

维普期刊数据库

Physical Layer Encryption of OFDM-PON Based on Quantum Noise Stream Cipher with Polar Code

在线全文

学校读者我要写书评

暂无评论

China Communications 2024年第3期21卷 174-188页

作者： Xu Yinbo Gao Mingyi Zhu Huaqing Chen Bowen Xiang Lian Shen Gangxiang Jiangsu engineering Research Center of Novel Optical Fiber Technology and Communication network Suzhou Key Laboratory of Advanced Optical Communication Network Technologythe School of Electronic and Information EngineeringSoochow UniversityNo.1 Shizi StreetSuzhou 215006China

Orthogonal frequency division multiplexing passive optical network(OFDM-PON) has superior anti-dispersion property to operate in the C-band of fiber for increased optical power budget. However,the downlink broadcast exposes the physical layer vulnerable to the threat of illegal eavesdropping. Quantum noise stream cipher(QNSC) is a classic physical layer encryption method and well compatible with the OFDM-PON. Meanwhile, it is indispensable to exploit forward error correction(FEC) to control errors in data transmission. However, when QNSC and FEC are jointly coded, the redundant information becomes heavier and thus the code rate of the transmitted signal will be largely reduced. In this work, we propose a physical layer encryption scheme based on polar-code-assisted QNSC. In order to improve the code rate and security of the transmitted signal, we exploit chaotic sequences to yield the redundant bits and utilize the redundant information of the polar code to generate the higher-order encrypted signal in the QNSC scheme with the operation of the *** experimentally demonstrate the encrypted 16/64-QAM, 16/256-QAM, 16/1024-QAM, 16/4096-QAM QNSC signals transmitted over 30-km standard single mode fiber. For the transmitted 16/4096-QAM QNSC signal, compared with the conventional QNSC method, the proposed method increases the code rate from 0.1 to 0.32 with enhanced security.

关键词： physical layer encryption polar code quantum noise stream cipher

维普期刊数据库

Resonant Beam Communications:A New Design Paradigm and Challenges

在线全文

学校读者我要写书评

暂无评论

Journal of Communications and information networks 2024年第1期9卷 80-87页

作者： Yuanming Tian Dongxu Li Chuan Huang Qingwen Liu Shengli Zhou Future network of Intelligence Institute and School of Science and engineering the Chinese University of Hong KongShenzhen 518172China School of Science and engineering and the Future network of Intelligence Institute the Chinese University of Hong KongShenzhen 518172China College of Electronics and information engineering Tongji UniversityShanghai 201804China department of Electrical and Computer engineering University of ConnecticutStorrsCT 06250USA

Resonant beam communications (RBCom), which adopts oscillating photons between two separate retroreflectors for information transmission, exhibits potential advantages over other types of wireless optical communications (WOC). However, echo interference generated by the modulated beam reflected from the receiver affects the transmission of the desired information. To tackle this challenge, a synchronization-based point-to-point RBCom system is proposed to eliminate the echo interference, and the design for the transmitter and receiver is discussed. Subsequently,the performance of the proposed RBCom is evaluated and compared with that of visible light communications(VLC)and free space optical communications (FOC). Finally, future research directions are outlined and several implementation challenges of RBCom systems are highlighted.

关键词： resonant beam communications(RBCom) wireless optical communications(WOC) echo interference elimination mobility multiple access

A quantum blind signature scheme based on dense coding for non-entangled states

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

Chinese Physics B 2024年第6期33卷 220-228页

作者：邢柯殷爱菡薛勇奇 department of information engineering East China Jiaotong UniversityNanchang 330013China

In some schemes, quantum blind signatures require the use of difficult-to-prepare multiparticle entangled states. By considering the communication overhead, quantum operation complexity, verification efficiency and other relevant factors in practical situations, this article proposes a non-entangled quantum blind signature scheme based on dense encoding. The information owner utilizes dense encoding and hash functions to blind the information while reducing the use of quantum resources. After receiving particles, the signer encrypts the message using a one-way function and performs a Hadamard gate operation on the selected single photon to generate the signature. Then the verifier performs a Hadamard gate inverse operation on the signature and combines it with the encoding rules to restore the message and complete the *** with some typical quantum blind signature protocols, this protocol has strong blindness in privacy protection,and higher flexibility in scalability and application. The signer can adjust the signature operation according to the actual situation, which greatly simplifies the complexity of the signature. By simultaneously utilizing the secondary distribution and rearrangement of non-entangled quantum states, a non-entangled quantum state representation of three bits of classical information is achieved, reducing the use of a large amount of quantum resources and lowering implementation costs. This improves both signature verification efficiency and communication efficiency while, at the same time, this scheme meets the requirements of unforgeability, non-repudiation, and prevention of information leakage.

关键词： quantum blind signature dense coding non-entanglement Hadamard gate

维普期刊数据库