检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Yang, Yuting Lei, Wenqiang Huang, Pei Cao, Juan Li, Jintao Chua, Tat-Seng University of Chinese Academy of Sciences Beijing China Sichuan University Sichuan China Stanford University California United States Key Lab of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China National University of Singapore Singapore

Dialogue state tracking (DST) module is an important component for task-oriented dialog systems to understand users' goals and needs. Collecting dialogue state labels including slots and values can be costly, especially with the wide application of dialogue systems in more and more new-rising domains. In this paper, we focus on how to utilize the language understanding and generation ability of pre-trained language models for DST. We design a dual prompt learning framework for few-shot DST. Specifically, we consider the learning of slot generation and value generation as dual tasks, and two prompts are designed based on such a dual structure to incorporate task-related knowledge of these two tasks respectively. In this way, the DST task can be formulated as a language modeling task efficiently under few-shot settings. Experimental results on two task-oriented dialogue datasets show that the proposed method not only outperforms existing state-of-the-art few-shot methods, but also can generate unseen slots. It indicates that DST-related knowledge can be probed from PLM and utilized to address low-resource DST efficiently with the help of prompt learning. Copyright © 2022, The Authors. All rights reserved.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

M2TR: Multi-modal Multi-scale Transformers for Deepfake Detection

arXiv

引用

arXiv 2021年

作者： Wang, Junke Wu, Zuxuan Ouyang, Wenhao Han, Xintong Chen, Jingjing Lim, Ser-Nam Jiang, Yu-Gang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University China Shanghai Collaborative Innovation Center on Intelligent Visual Computing China Huya Inc Meta AI

The widespread dissemination of Deepfakes demands effective approaches that can detect perceptually convincing forged images. In this paper, we aim to capture the subtle manipulation artifacts at different scales using transformer models. In particular, we introduce a Multi-modal Multi-scale TRansformer (M2TR), which operates on patches of different sizes to detect local inconsistencies in images at different spatial levels. M2TR further learns to detect forgery artifacts in the frequency domain to complement RGB information through a carefully designed cross modality fusion block. In addition, to stimulate Deepfake detection research, we introduce a high-quality Deepfake dataset, SR-DF, which consists of 4,000 DeepFake videos generated by state-of-the-art face swapping and facial reenactment methods. We conduct extensive experiments to verify the effectiveness of the proposed method, which outperforms state-of-the-art Deepfake detection methods by clear margins. Copyright © 2021, The Authors. All rights reserved.

关键词： Frequency domain analysis

来源：评论

学校读者我要写书评

暂无评论

MMDSSE: Multi-client and Multi-keyword Dynamic Searchable Symmetric Encryption for Cloud Storage

MMDSSE: Multi-client and Multi-keyword Dynamic Searchable Sy...

引用

Annual Conference on Privacy, Security and Trust, PST

作者： Panyu Wu Zhenfu Cao Jiachen Shen Xiaolei Dong Yihao Yang Jun Zhou Liming Fang Zhe Liu Chunpeng Ge Chunhua Su Shanghai Key Laboratory of Trustworthy Computing East China Normal University Shanghai China Research Center for Basic Theories of Intelligent Computing Research Institute of Basic Theories Zhejiang Lab Hangzhou China College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China Science and Technology on Parallel and Distributed Processing Laboratory (PDL) Changsha China Shandong University Jinan China University of Aizu Fukushima Japan

Since data outsourcing poses privacy concerns with data leakage, searchable symmetric encryption (SSE) has emerged as a powerful solution that enables clients to perform query operations on encrypted data while preserving their privacy. Dynamic SSE schemes have been proposed to handle update operations. However, it is shown that updates might increase the risk of information leakage. Meanwhile, to meet the requirement of real-world applications, it is desirable to have the searchable encryption scheme which supports both multiple clients and multi-keyword queries. To address these issues, this paper proposes MMDSSE, a multi-client forward secure dynamic SSE scheme that supports multi-keyword queries. MMDSSE allows the clients narrow down the results by providing an arbitrary subset of the entire archive, and thus suitable for cloud storage environment. Security analysis and experimental evaluations show that MMDSSE is secure and efficient.

关键词：

来源：评论

学校读者我要写书评

暂无评论

research on Highly Anti-Humidity and Selective NO2 Detection Based on GaN/MoO3 N-N Heterojunction

SSRN

引用

SSRN 2024年

作者： Hong, Yutao Han, Dan Li, Donghui He, Xiuli Zhao, Li Wang, Weidong Li, Hongwei Sang, Shengbo Liang, Hua Shanxi Key Laboratory of Micro Nano Sensors & Artificial Intelligence Perception College of Electronic Information and Optical Engineering Taiyuan University of Technology Taiyuan030024 China Key Lab of Advanced Transducers and Intelligent Control System Ministry of Education Taiyuan University of Technology Taiyuan030024 China State Key Laboratory of Transducer Technology Aerospace Information Research Institute Chinese Academy of Sciences Beijing100190 China Bioengineering Research Center Medical Innovation Research Division Chinese PLA General Hospital Beijing China Shanxi Academy of Medical Sciences China

Effective and accuratedetection of NO2 gas is crucial for protecting the environment and public health. In this work, balsam pear-like GaN/MoO3 porous nanosheets (GaN/MoO3) composites were synthesized by the in situ assembling, the hydrothermal method, and the nitridation process to detect NO2. A series of characterization methods were used to manifest the successful synthesis and to analyze the morphology and elemental composition features of GaN/MoO3 composites. The results of the gas-sensing experiments show that at the operating temperature of 225 °C, the gas sensor based on the optimal ratio of GaN/MoO3 composites exhibits a low theoretical limit of detection (0.78 ppb) for NO2, which is significantly improved compared with the limit of detection of MoO3 (5 ppm). Further, the response value of the sensor based on GaN/MoO3 composites to 200 ppm NO2 (44.74) is 4.36 times higher than that of the sensor based on pure GaN, with an improvement in stability. Meanwhile, the sensor also possessed excellent repeatability, selectivity, stability, and splendid anti-humidity ability at 15-70 % RH. The improved gas-sensing performance is ascribed to the unique microstructure of both GaN and MoO3, and the heterojunction effect between them. This work provides a way to improve the efficiency and anti-humidity ability of real-time monitoring of NO2 for the sensors of MoO3-based composite. © 2024, The Authors. All rights reserved.

关键词： Gallium nitride

来源：评论

学校读者我要写书评

暂无评论

A Prompting-based Approach for Adversarial Example Generation and Robustness Enhancement

arXiv

引用

arXiv 2022年

作者： Yang, Yuting Huang, Pei Cao, Juan Li, Jintao Lin, Yun Dong, Jin Song Ma, Feifei Zhang, Jian Key Lab of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Beijing China National University of Singapore Singapore Laboratory of Parallel Software and Computational Science ISCAS Beijing China

Recent years have seen the wide application of NLP models in crucial areas such as finance, medical treatment, and news media, raising concerns of the model robustness and vulnerabilities. In this paper, we propose a novel prompt-based adversarial attack to compromise NLP models and robustness enhancement technique. We first construct malicious prompts for each instance and generate adversarial examples via mask-and-filling under the effect of a malicious purpose. Our attack technique targets the inherent vulnerabilities of NLP models, allowing us to generate samples even without interacting with the victim NLP model, as long as it is based on pre-trained language models (PLMs). Furthermore, we design a prompt-based adversarial training method to improve the robustness of PLMs. As our training method does not actually generate adversarial samples, it can be applied to large-scale training sets efficiently. The experimental results show that our attack method can achieve a high attack success rate with more diverse, fluent and natural adversarial examples. In addition, our robustness enhancement method can significantly improve the robustness of models to resist adversarial attacks. Our work indicates that prompting paradigm has great potential in probing some fundamental flaws of PLMs and fine-tuning them for downstream tasks. Copyright © 2022, The Authors. All rights reserved.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Context-aware Multi-level Question Embedding Fusion for visual question answering

引用

information Fusion 2024年 102卷

作者： Li, Shengdong Gong, Chen Zhu, Yuqing Luo, Chuanwen Hong, Yi Lv, Xueqiang School of Information Renmin University of China Beijing100872 China School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China Zhejiang Lab Zhejiang Hangzhou311121 China Department of Computer Science California State University Los Angeles CA90032 United States School of Information Science and Technology Beijing Forestry University Beijing100083 China Engineering Research Center for Forestry-Oriented Intelligent Information Processing National Forestry and Grassland Administration 100083 China Beijing Key Laboratory of Internet Culture and Digital Dissemination Research Beijing Information Science and Technology University 100101 China

Question model has been widely concerned as the cornerstone of constructing Visual Question Answering (VQA) models. Existing question models attempt to exploit word context to extract multi-level concepts for modeling multi-level questions. However, they still have many defects. For example, most question models utilize simple fusion methods to fuse shallow modules and extract parameter-unshared low-level concepts, leading to poor modeling of multi-level questions;although some question models use deep bidirectional Transformer encoder in external knowledge transfer and BERT for multi-level questions, their complexity is still high. To solve these issues, we propose a novel low-complex multi-level contextual question model, termed Context-aware Multi-level Question Embedding Fusion (CMQEF). We formalize its concepts and theories, deduce its modeling process, optimization process and feature extraction process, analyze its low complexity and high expressiveness, and prove that it defines a new way to solve parameter non-sharing for extracting parameter-shared multi-level concepts and optimize the tradeoff between expressiveness and complexity in question models. Extensive experiments on VQAv2 and VQA-CPv2 validate that comparing with the state-of-the-art, our CMQEF outperforms it on SANs and UpDn, reduces the language priors of SANs and UpDn, and has preferable interpretability and applicability. Our code is available at https://***/lsdruc/CMQEF-for-VQA. © 2023 The Author(s)

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Self-Mutual Distillation Learning for Continuous Sign Language Recognition

Self-Mutual Distillation Learning for Continuous Sign Langua...

引用

International Conference on Computer Vision (ICCV)

作者： Aiming Hao Yuecong Min Xilin Chen Key Lab of Intelligent Information Processing of Chinese Academy of Sciences (CAS) Institute of Computing Technology CAS Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781665428132

In recent years, deep learning moves video-based Continuous Sign Language Recognition (CSLR) significantly forward. Currently, a typical network combination for CSLR includes a visual module, which focuses on spatial and short-temporal information, followed by a contextual module, which focuses on long-temporal information, and the Connectionist Temporal Classification (CTC) loss is adopted to train the network. However, due to the limitation of chain rules in back-propagation, the visual module is hard to adjust for seeking optimized visual features. As a result, it enforces that the contextual module focuses on contextual information optimization only rather than balancing efficient visual and contextual information. In this paper, we propose a Self-Mutual Knowledge Distillation (SMKD) method, which enforces the visual and contextual modules to focus on short-term and long-term information and enhances the discriminative power of both modules simultaneously. Specifically, the visual and contextual modules share the weights of their corresponding classifiers, and train with CTC loss simultaneously. Moreover, the spike phenomenon widely exists with CTC loss. Although it can help us choose a few of the key frames of a gloss, it does drop other frames in a gloss and makes the visual feature saturation in the early stage. A gloss segmentation is developed to relieve the spike phenomenon and decrease saturation in the visual module. We conduct experiments on two CSLR bench-marks: PHOENIX14 and PHOENIX14-T. Experimental results demonstrate the effectiveness of the SMKD.

关键词： Training Deep learning Visualization Computer vision Gesture recognition Assistive technologies Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

Cross-domain Contrastive Learning for Unsupervised Domain Adaptation

arXiv

引用

arXiv 2021年

作者： Wang, Rui Wu, Zuxuan Weng, Zejia Chen, Jingjing Qi, Guo-Jun Jiang, Yu-Gang Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai Collaborative Innovation Center on Intelligent Visual Computing China Seattle Cloud Lab Futurewei Technologies BellevueWA98004 United States

Unsupervised domain adaptation (UDA) aims to transfer knowledge learned from a fully-labeled source domain to a different unlabeled target domain. Most existing UDA methods learn domain-invariant feature representations by minimizing feature distances across domains. In this work, we build upon contrastive self-supervised learning to align features so as to reduce the domain discrepancy between training and testing sets. Exploring the same set of categories shared by both domains, we introduce a simple yet effective framework CDCL, for domain alignment. In particular, given an anchor image from one domain, we minimize its distances to cross-domain samples from the same class relative to those from different categories. Since target labels are unavailable, we use a clustering-based approach with carefully initialized centers to produce pseudo labels. In addition, we demonstrate that CDCL is a general framework and can be adapted to the data-free setting, where the source data are unavailable during training, with minimal modification. We conduct experiments on two widely used domain adaptation benchmarks, i.e., Office-31 and VisDA-2017, for image classification tasks, and demonstrate that CDCL achieves state-of-the-art performance on both datasets. Copyright © 2021, The Authors. All rights reserved.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Low-Noise Microwave Signal Generation by Heterodyning Two Narrow-Linewidth Self-Injection Locked Lasers

Low-Noise Microwave Signal Generation by Heterodyning Two Na...

引用

Conference on Lasers and Electro-Optics (CLEO)

作者： Yihao Fan Siyu E Yuyao Guo Xinhang Li Liangjun Lu Shuxiao Wang Yan Cai Jianping Chen Linjie Zhou Department of Electronic Engineering State Key Laboratory of Advanced Optical Communication Systems and Networks Shanghai Key Lab of Navigation and Location Services Shanghai Jiao Tong University Shanghai China SJTU-Pinghu Institute of Intelligent Optoelectronics Pinghu China State Key Laboratory of Functional Materials for Informatics Shanghai Institute of Microsystem and Information Technology Chinese Academy of Sciences Shanghai China Shanghai Industrial μ Technology Research Institute Shanghai China

We realize a hybrid integrated self-injection locking laser (SIL) with an ultra-narrow intrinsic linewidth of 5 Hz. With heterodyne synthesis using a pair of SILs, a microwave signal with a 4-kHz linewidth is achieved.

ISBN: (纸本)9798350369311

关键词： Wireless communication Masers Microwave communication Microwave theory and techniques Hybrid power systems Microwave generation Microwave photonics Laser applications Electro-optical waveguides

来源：评论

学校读者我要写书评

暂无评论

A Novel Magnetoelastic Biosensor Based on Pdms/Fesib/Qds Composite Film for African Swine Fever Virus Detections

SSRN

引用

SSRN 2024年

作者： Liu, Yuanhang Sang, Shengbo Zhao, Dong Ge, Yang Xue, Juanjuan Duan, Qianqian Guo, Xing Shanxi Key Laboratory of Micro Nano Sensors & Artificial Intelligence Perception College of Electronic Information and Optical Engineering Taiyuan University of Technology Taiyuan030024 China Key Lab of Advanced Transducers and Intelligent Control System of The Ministry of Education Taiyuan University of Technology Taiyuan030024 China Shanxi Research Institute of 6D Artificial Intelligence Biomedical Science Taiyuan030024 China

African swine fever (ASF) is a highly contagious and severe hemorrhagic disease caused by the African swine fever virus (ASFV). The continuous spread of ASFV affects the safety of global meat supply, so the establishment of sensitive and specific detection methods for ASFV has become an important hot spot in food safety. Herein, we developed a flexible magnetoelastic (ME) biosensor based on PDMS/FeSiB/QDs composite films for the detection of ASFV. Based on the high luminescence performance of CsPbBr3 quantum dots and the excellent magnetoelastic effect of FeSiB, flexible ME biosensors convert stress signals generated by antibody-antigen-specific binding into optical and electromagnetic signals. The nanostructures covalently linked by quantum dots and PDMS provide biomodification sites for ASFV antibodies, simplifying the functionalization modification process compared to conventional biosensors. The deformation of the PDMS film is amplified and conversion of surface stress signal to electrical signal is enhanced by exposing the biosensor to a uniform magnetic field. The experimental results proved that the flexible ME biosensor has wide linear range of 10 ng/mL - 100 μg/mL, and the detection limit is as low as 0.7933 ng/mL. Moreover, the flexible ME biosensor have also shown to good stability, sensitivity and specificity. © 2024, The Authors. All rights reserved.

关键词： Biosensors

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：