检索结果-内蒙古大学图书馆

Multi-scale Contrastive Learning for Gastroenteroscopy Classification

学校读者我要写书评

暂无评论

Multi-scale Contrastive Learning for Gastroenteroscopy Class...

Annual IEEE Symposium on Computer-Based Medical systems

作者： Dan Li Xuechen Li Zhibin Peng Wenting Chen Linlin Shen Guangyao Wu Computer Vision Institute School of Computer Science & Software Engineering Shenzhen University National Engineering Laboratory for Big Data System Computing Technology ShenZhen University Shenzhen China City University of Hong Kong Hong Kong SAR China Shenzhen Institute of Artificial Intelligence & Robotics for Society Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University General Hospital

In gastroenteroscopy image analysis, numerous CADs demonstrate that deep learning aids doctors' diagnosis. The shapes and sizes of the lesions are varied. And in the clinic, the dataset appears to be data imbalanced. However, existing methods directly classify by texture and ignore lesions with various shapes and sizes. To address the issue above, we propose a deep neural network, which consists of multi-scale feature extraction, contrastive feature learning and a multi-scale feature fusion module. We train the contrastive feature learning module and multi-scale feature fusion module simultaneously to alleviate the issue of data distribution differences. Thus, the proposed network can better identify various categories. Extensive experiments on the Hyper Kvasir dataset show that the proposed Hybrid-M2CL outperforms the benchmark proposed by the dataset with 5.0% Macro Precision, 3.3% Macro Recall, 3.4% Macro F1-score, 3.3% Micro Precision, 3.6% MCC. In addition, it outperforms the SOTA by 1.1% Macro F1-score, 2.6% MCC, and 2.0% B-ACC.

关键词：

ReconBoost: boosting can achieve modality reconcilement 24

学校读者我要写书评

暂无评论

ReconBoost: boosting can achieve modality reconcilement

Proceedings of the 41st International Conference on Machine Learning

作者： Cong Hua Qianqian Xu Shilong Bao Zhiyong Yang Qingming Huang Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China and School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China Institute of Information Engineering Chinese Academy of Sciences Beijing China and School of Cyber Security University of Chinese Academy of Sciences Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China and Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China and Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of Sciences Beijing China

This paper explores a novel multi-modal alternating learning paradigm pursuing a reconciliation between the exploitation of uni-modal features and the exploration of cross-modal interactions. This is motivated by the fact that current paradigms of multi-modal learning tend to explore multi-modal features simultaneously. The resulting gradient prohibits further exploitation of the features in the weak modality, leading to modality competition, where the dominant modality overpowers the learning process. To address this issue, we study the modality-alternating learning paradigm to achieve reconcilement. Specifically, we propose a new method called ReconBoost to update a fixed modality each time. Herein, the learning objective is dynamically adjusted with a reconcilement regularization against competition with the historical models. By choosing a KL-based reconcilement, we show that the proposed method resembles Friedman's Gradient-Boosting (GB) algorithm, where the updated learner can correct errors made by others and help enhance the overall performance. The major difference with the classic GB is that we only preserve the newest model for each modality to avoid overfitting caused by ensembling strong learners. Furthermore, we propose a memory consolidation scheme and a global rectification scheme to make this strategy more effective. Experiments over six multi-modal benchmarks speak to the efficacy of the method. We release the code at https://***/huacong/ReconBoost.

关键词：

Quantum reversible circuits for multiplicative inverse

学校读者我要写书评

暂无评论

EPJ Quantum Technology 2022年第1期9卷 1-17页

作者： Luo, Qing-bin Yang, Guo-wu Li, Xiao-yu Li, Qiang Department of Computer Science and Technology School of Information Engineering Hubei Minzu University Enshi China School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu China Big data research Center & School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China Advanced Cryptography and System Security Key Laboratory of Sichuan Province Chengdu University of Information Technology Chengdu China

The synthesis of quantum circuits for multiplicative inverse over $\operatorname{GF}(2^{8})$ are discussed in this paper. We first convert the multiplicative inverse operation in $\operatorname{GF}(2^{8})$ to arithmetic operations in the composite field $\operatorname{GF}((2^{4})^{2})$ , and then discuss the expressions of the square calculation, the inversion calculation and the multiplication calculation separately in the finite field $\operatorname{GF}(2^{4})$ , where the expressions of multiplication calculation in $\operatorname{GF}(2^{4})$ are given directly in $\operatorname{GF}(2^{4})$ and given through being transformed into the composite field $\operatorname{GF}((2^{2})^{2})$ . Then the quantum circuits of these calculations are realized one by one. Finally, two quantum circuits for multiplicative inverse over $\operatorname{GF}(2^{8})$ are synthesized. They both use 21 qubits, the first quantum circuit uses 55 Toffoli gates and 107 CNOT gates and the second one uses 37 Toffoli gates and 209 CNOT gates. As an example of the application of multiplication inverse, we apply these quantum circuits to the implementations of the S-box quantum circuit of the AES cryptographic algorithm. Two quantum circuits for implementing the S-box of the AES cryptographic algorithm are presented. The first quantum circuit uses 21 qubits, 55 Toffoli gates, 131 CNOT gates and 4 NOT gates and the second one uses 21 qubits, 37 Toffoli gates, 233 CNOT gates and 4 NOT gates. Through the evaluation of quantum cost, the two quantum circuits of the S-box of AES cryptographic algorithm use less quantum resources than the existing schemes.

关键词：

A Software Defect Prediction Method Based on Learnable Three-Line Hybrid Feature Fusion

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Tang, Yu Dai, Qi Du, Ye Chen, Lifang Niu, Xuanwen School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China Department of Automation College of Information Science and Engineering China University of Petroleum-Beijing Beijing China Beijing Key Laboratory of Security and Privacy in Intelligent Transportation Beijing Jiaotong University Beijing100044 China College of Science North China University of Science and Technology Hebei Tangshan063210 China Hebei Key Laboratory of Data Science and Application Hebei Tangshan063210 China

Software defect prediction plays a crucial role in ensuring the security and quality of software systems. However, it faces challenges posed by high-dimensional features present in software defect datasets and the limited effectiveness of traditional nonlinear dimensionality reduction methods in extracting essential feature information. To address these issues, we propose a novel approach called learnable three-line hybrid feature fusion (LTHFFA), which incorporates the principle of three-line hybrid breeding into feature fusion for the first time. In this method, three distinct dimensionality reduction techniques are initially employed to obtain three separate sets of features. Subsequently, a learnable weight factor feature fusion method is proposed to facilitate automatic learning and dynamic adjustment of feature weights. By integrating the three feature sets based on the principle of three-line hybrid breeding, we derive learnable three-line hybrid fusion features. These features are then utilized in the context of software defect prediction. Experimental results demonstrate the superior performance of LTHFFA compared to nine other dimensionality reduction methods across seventeen publicly available software defect datasets. LTHFFA exhibits the ability to effectively integrate multiple feature sets, reduce feature redundancy, and enhance predictive accuracy. Moreover, statistical analysis using Friedman ranking and Holm's post-hoc test confirms the significant advantage of LTHFFA over alternative dimensionality reduction methods. © 2023, The Authors. All rights reserved.

关键词： Defects

Disentangled Noisy Correspondence Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Dang, Zhuohang Luo, Minnan Wang, Jihong Jia, Chengyou Han, Haochen Wan, Herun Dai, Guang Chang, Xiaojun Wang, Jingdong The School of Computer Science and Technology The Ministry of Education Key Laboratory of Intelligent Networks and Network Security The Shaanxi Province Key Laboratory of Big Data Knowledge Engineering Xi'an Jiaotong University Shaanxi Xi'An710049 China SGIT AI Lab State Grid Shaanxi Electric Power Company Limited State Grid Corporation of China Shaanxi China The School of Information Science and Technology University of Science and Technology China United Arab Emirates The Baidu Inc China

Cross-modal retrieval is crucial in understanding latent correspondences across modalities. However, existing methods implicitly assume well-matched training data, which is impractical as real-world data inevitably involves imperfect alignments, i.e., noisy correspondences. Although some works explore similarity-based strategies to address such noise, they suffer from sub-optimal similarity predictions influenced by modality-exclusive information (MEI), e.g., background noise in images and abstract definitions in texts. This issue arises as MEI is not shared across modalities, thus aligning it in training can markedly mislead similarity predictions. Moreover, although intuitive, directly applying previous cross-modal disentanglement methods suffers from limited noise tolerance and disentanglement efficacy. Inspired by the robustness of information bottlenecks against noise, we introduce DisNCL, a novel information-theoretic framework for feature Disentanglement in Noisy Correspondence Learning, to adaptively balance the extraction of MII and MEI with certifiable optimal cross-modal disentanglement efficacy. DisNCL then enhances similarity predictions in modality-invariant subspace, thereby greatly boosting similarity-based alleviation strategy for noisy correspondences. Furthermore, DisNCL introduces soft matching targets to model noisy many-to-many relationships inherent in multi-modal input for noise-robust and accurate cross-modal alignment. Extensive experiments confirm DisNCL's efficacy by 2% average recall improvement. Mutual information estimation and visualization results show that DisNCL learns meaningful MII/MEI subspaces, validating our theoretical analyses. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Meng, Benyuan Xu, Qianqian Wang, Zitai Cao, Xiaochun Huang, Qingming Institute of Information Engineering CAS China School of Cyber Security University of Chinese Academy of Sciences China Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS China Peng Cheng Laboratory China School of Cyber Science and Tech. Shenzhen Campus of Sun Yat-sen University China School of Computer Science and Tech. University of Chinese Academy of Sciences China Key Laboratory of Big Data Mining and Knowledge Management CAS China

Diffusion models are initially designed for image generation. Recent research shows that the internal signals within their backbones, named activations, can also serve as dense features for various discriminative tasks such as semantic segmentation. Given numerous activations, selecting a small yet effective subset poses a fundamental problem. To this end, the early study of this field performs a large-scale quantitative comparison of the discriminative ability of the activations. However, we find that many potential activations have not been evaluated, such as the queries and keys used to compute attention scores. Moreover, recent advancements in diffusion architectures bring many new activations, such as those within embedded ViT modules. Both combined, activation selection remains unresolved but overlooked. To tackle this issue, this paper takes a further step with a much broader range of activations evaluated. Considering the significant increase in activations, a full-scale quantitative comparison is no longer operational. Instead, we seek to understand the properties of these activations, such that the activations that are clearly inferior can be filtered out in advance via simple qualitative evaluation. After careful analysis, we discover three properties universal among diffusion models, enabling this study to go beyond specific models. On top of this, we present effective feature selection solutions for several popular diffusion models. Finally, the experiments across multiple discriminative tasks validate the superiority of our method over the SOTA competitors. Our code is available at this url. © 2024, CC BY.

关键词： Semantics

PrintListener: Uncovering the Vulnerability of Fingerprint Authentication via the Finger Friction Sound

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhou, Man Su, Shuao Wang, Qian Li, Qi Zhou, Yuting Ma, Xiaojing Li, Zhengxiong School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Cyber Science and Engineering Wuhan University China Institute for Network Sciences and Cyberspace Tsinghua University China Department of Computer Science and Engineering University of Colorado Denver United States Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Huazhong University of Science and Technology China

Fingerprint authentication has been extensively employed in contemporary identity verification systems owing to its rapidity and cost-effectiveness. Due to its widespread use, fingerprint leakage may cause sensitive information theft, enormous economic and personnel losses, and even a potential compromise of national security. As a fingerprint that can coincidentally match a specific proportion of the overall fingerprint population, MasterPrint rings the alarm bells for the security of fingerprint authentication. In this paper, we propose a new side-channel attack on the minutiae-based Automatic Fingerprint Identification system (AFIS), called PrintListener, which leverages users’ fingertip swiping actions on the screen to extract fingerprint pattern features (the first-level features) and synthesizes a stronger targeted PatternMasterPrint with potential second-level features. The attack scenario of PrintListener is extensive and covert. It only needs to record users’ fingertip friction sound and can be launched by leveraging a large number of social media platforms. Extensive experimental results in real-world scenarios show that Printlistener can significantly improve the attack potency of MasterPrint. © 2024, CC0.

关键词： Friction

MFCLIP: Multi-modal Fine-grained CLIP for Generalizable Diffusion Face Forgery Detection

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Yaning Wang, Tianyi Yu, Zitong Gao, Zan Shen, Linlin Chen, Shengyong Jinan250353 China Nanyang Technological University 50 Nanyang Ave Block N 4 639798 Singapore School of Computing and Information Technology Great Bay University Dongguan523000 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen518060 China Jinan250014 China The Key Laboratory of Computer Vision and System Ministry of Education Tianjin University of Technology Tianjin300384 China Computer Vision Institute College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen518129 China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen518060 China

The rapid development of photo-realistic face generation methods has raised significant concerns in society and academia, highlighting the urgent need for robust and generalizable face forgery detection (FFD) techniques. Although existing approaches mainly capture face forgery patterns using image modality, other modalities like fine-grained noises and texts are not fully explored, which limits the generalization capability of the model. In addition, most FFD methods tend to identify facial images generated by GAN, but struggle to detect unseen diffusion-synthesized ones. To address the limitations, we aim to leverage the cutting-edge foundation model, contrastive language-image pre-training (CLIP), to achieve generalizable diffusion face forgery detection (DFFD). In this paper, we propose a novel multi-modal fine-grained CLIP (MFCLIP) model, which mines comprehensive and fine-grained forgery traces across image-noise modalities via language-guided face forgery representation learning, to facilitate the advancement of DFFD. Specifically, we devise a fine-grained language encoder (FLE) that extracts fine global language features from hierarchical text prompts. We design a multi-modal vision encoder (MVE) to capture global image forgery embeddings as well as fine-grained noise forgery patterns extracted from the richest patch, and integrate them to mine general visual forgery traces. Moreover, we build an innovative plug-and-play sample pair attention (SPA) method to emphasize relevant negative pairs and suppress irrelevant ones, allowing cross-modality sample pairs to conduct more flexible alignment. Extensive experiments and visualizations show that our model outperforms the state of the arts on different settings like cross-generator, cross-forgery, and cross-dataset evaluations. Copyright © 2024, The Authors. All rights reserved.

关键词： Alignment

data-Driven Compressed Sensing for Massive Wireless Access

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Bai, Yanna Chen, Wei Sun, Feifei Ai, Bo Popovski, Petar The State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University Beijing100044 China Frontiers Science Center for Smart High-Speed Railway System Beijing100044 China Beijing China ResearchCenter of Networks and Communications Peng Cheng Laboratory Shenzhen518055 China Henan Joint International Research Laboratory of Intelligent Networking and Data Analysis Zhengzhou University Zhengzhou450001 China Aalborg University Denmark

The central challenge in massive machine-type communications (mMTC) is to connect a large number of uncoordinated devices through a limited spectrum. The typical mMTC communication pattern is sporadic, with short packets. This could be exploited in grant-free random access in which the activity detection, channel estimation, and data recovery are formulated as a sparse recovery problem and solved via compressed sensing algorithms. This approach results in new challenges in terms of high computational complexity and latency. We present how data-driven methods can be applied in grant-free random access and demonstrate the performance gains. Variations of neural networks for the problem are discussed, as well as future challenges and potential directions. Copyright © 2022, The Authors. All rights reserved.

关键词： Compressed sensing