检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Song, Ziying Jia, Feiyang Pan, Hongyu Luo, Yadan Jia, Caiyan Zhang, Guoxin Liu, Lin Ji, Yang Yang, Lei Wang, Li School of Computer Science and Technology Beijing Jiaotong University China Beijing Key Lab of Traffic Data Analysis and Mining China Horizon Robotics The University of Queensland Australia Hebei University of Science and Technology China Tsinghua University China Beijing Institute of Technology China

In the field of 3D object detection tasks, fusing heterogeneous features from LiDAR and camera sensors into a unified Bird's Eye View (BEV) representation is a widely adopted paradigm. However, existing methods are often compromised by imprecise sensor calibration, resulting in feature misalignment in LiDAR-camera BEV fusion. Moreover, such inaccuracies result in errors in depth estimation for the camera branch, ultimately causing misalignment between LiDAR and camera BEV features. In this work, we propose a novel ContrastAlign approach that utilizes contrastive learning to enhance the alignment of heterogeneous modalities, thereby improving the robustness of the fusion process. Specifically, our approach includes the L-Instance module, which directly outputs LiDAR instance features within LiDAR BEV features. Then, we introduce the C-Instance module, which predicts camera instance features through RoI (Region of Interest) pooling on the camera BEV features. We propose the InstanceFusion module, which utilizes contrastive learning to generate similar instance features across heterogeneous modalities. We then use graph matching to calculate the similarity between the neighboring camera instance features and the similarity instance features to complete the alignment of instance features. Our method achieves state-of-the-art performance, with an mAP of 70.3%, surpassing BEVFusion by 1.8% on the nuScenes validation set. Importantly, our method outperforms BEVFusion by 7.3% under conditions with misalignment noise. Copyright © 2024, The Authors. All rights reserved.

关键词： Cameras

来源：评论

学校读者我要写书评

暂无评论

Strong Transferable Adversarial Attacks via Ensembled Asymptotically Normal Distribution Learning

Strong Transferable Adversarial Attacks via Ensembled Asympt...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zhengwei Fang Rui Wang Tao Huang Liping Jing School of Computer Science and Technology Beijing Jiaotong University Beijing China Beijing Key Lab of Traffic Data Analysis and Mining Beijing China Collaborative Innovation Center of Railway Traffic Safety Beijing China

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Strong adversarial examples are crucial for evaluating and enhancing the robustness of deep neural networks. However, the performance of popular attacks is usually sensitive, for instance, to minor image transformations, stemming from limited information ─ typically only one input example, a handful of white-box source models, and undefined defense strategies. Hence, the crafted adversarial examples are prone to overfit the source model, which hampers their transferability to unknown architectures. In this paper, we propose an approach named Multiple Asymptotically Normal Distribution Attacks (MultiANDA) which explicitly characterize adversarial perturbations from a learned distribution. Specifically, we approximate the posterior distribution over the perturbations by taking advantage of the asymptotic normality property of stochastic gradient ascent (SGA), then employ the deep ensemble strategy as an effective proxy for Bayesian marginalization in this process, aiming to estimate a mixture of Gaussians that facilitates a more thorough exploration of the potential optimization space. The approximated posterior essentially describes the stationary distribution of SGA iterations, which captures the geometric information around the local optimum. Thus, MultiANDA allows drawing an unlimited number of adversarial perturbations for each input and reliably maintains the transferability. Our proposed method outperforms ten state-of-the-art black-box attacks on deep learning models with or without defenses through extensive experiments on seven normally trained and seven defense models.

关键词： Deep learning Image transformation Perturbation methods Computational modeling Stochastic processes Gaussian distribution Robustness

来源：评论

学校读者我要写书评

暂无评论

SCALING UP MASKED DIFFUSION MODELS ON TEXT

arXiv

引用

arXiv 2024年

作者： Nie, Shen Zhu, Fengqi Du, Chao Pang, Tianyu Liu, Qian Zeng, Guangtao Lin, Min Li, Chongxuan Gaoling School of Artificial Intelligence Renmin University of China China Beijing Key Laboratory of Big Data Management and Analysis Methods China Sea AI Lab Singapore Singapore University of Technology and Design Singapore

Masked diffusion models (MDMs) have shown promise in language modeling, yet their scalability and effectiveness in core language tasks, such as text generation and language understanding, remain underexplored. This paper establishes the first scaling law for MDMs, demonstrating a scaling rate comparable to autoregressive models (ARMs) and a relatively small compute gap. Motivated by their scalability, we train a family of MDMs with up to 1.1 billion (B) parameters to systematically evaluate their performance against ARMs of comparable or larger sizes. Fully leveraging the probabilistic formulation of MDMs, we propose a simple yet effective unsupervised classifier-free guidance that effectively exploits large-scale unpaired data, boosting performance for conditional inference. In language understanding, the 1.1B MDM outperforms the 1.1B TinyLlama model trained on the same data across four of eight zero-shot benchmarks. Notably, it achieves competitive math reasoning ability with the 7B Llama-2 model on the GSM8K dataset. In text generation, MDMs with 16 times more pre-training time offer a flexible trade-off against ARMs with the accelerated sampling technique KV-Cache: MDMs match ARMs in performance while being 1.4 times faster during sampling. Moreover, MDMs address challenging tasks for ARMs by effectively handling bidirectional reasoning and adapting to temporal shifts in data. Notably, a 1.1B MDM breaks the reverse curse encountered by much larger ARMs with significantly more data and computation, such as 13B Llama-2 and 175B GPT-3. Our code is available at https://***/ML-GSAI/SMDM. Copyright © 2024, The Authors. All rights reserved.

关键词： Scalability

来源：评论

学校读者我要写书评

暂无评论

Multi-modal Multi-emotion Emotional Support Conversation 9th

Multi-modal Multi-emotion Emotional Support Conversation

引用

19th International Conference on Advanced data Mining and Applications, ADMA 2023

作者： Liu, Guangya Dong, Xiao Wang, Meng-xiang Yu, Jianxing Gan, Mengjiao Liu, Wei Yin, Jian School of Artificial Intelligence Sun Yat-sen University Zhuhai China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Pazhou Lab Guangzhou510330 China Key Laboratory of Sustainable Tourism Smart Assessment Technology Ministry of Culture and Tourism Zhuhai China China National Institute of Standardization Beijing100088 China

ISBN: (纸本)9783031466601

This paper proposes a new task of Multi-modal Multi-emotion Emotional Support Conversation (MMESC), which has great value in various applications, such as counseling, daily chatting, and elderly company. This task aims to fully perceive the users’ emotional states from multiple modalities and generate appropriate responses to provide comfort for improving their feelings. Traditional works mainly focus on textual conversation, while a single-modal cannot accurately reflect the users’ emotions, such as saying fine with an inconsistent disgusting feeling. To address this problem, we propose a new task on multi-modalities and exploit a new method called FEAT for this new task. FEAT can integrate fine-grained emotional knowledge from multiple modalities. It first recognizes the users’ mental states based on an emotion-aware transformer. It then generates supportive responses using a hybrid method with multiple comfort strategies. To evaluate our method, we construct a large-scale dataset named MMESConv. It is almost two times larger than existing single-modal datasets. There are three modalities in this dataset (text, audio, and video) with fine-grained emotion annotations and strategy labels. Extensive experiments on this dataset demonstrate the advantages of our proposed framework. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Behavioral research

来源：评论

学校读者我要写书评

暂无评论

Exploring the Privacy Protection Capabilities of Chinese Large Language Models

arXiv

引用

arXiv 2024年

作者： Yang, Yuqi Huang, Xiaowen Sang, Jitao School of Computer Science and Technology Beijing Jiaotong University Beijing China School of Computer Science and Technology Beijing Jiaotong University Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University Key Laboratory of Big Data Artificial Intelligence in Transportation Beijing Jiaotong University Ministry of Education Beijing China

Large language models (LLMs), renowned for their impressive capabilities in various tasks, have significantly advanced artificial intelligence. Yet, these advancements have raised growing concerns about privacy and security implications. To address these issues and explain the risks inherent in these models, we have devised a three-tiered progressive framework tailored for evaluating privacy in language systems. This framework consists of progressively complex and in-depth privacy test tasks at each tier. Our primary objective is to comprehensively evaluate the sensitivity of large language models to private information, examining how effectively they discern, manage, and safeguard sensitive data in diverse scenarios. This systematic evaluation helps us understand the degree to which these models comply with privacy protection guidelines and the effectiveness of their inherent safeguards against privacy breaches. Our observations indicate that existing Chinese large language models universally show privacy protection shortcomings. It seems that at the moment this widespread issue is unavoidable and may pose corresponding privacy risks in applications based on these models. © 2024, CC BY.

关键词： Sensitive data

来源：评论

学校读者我要写书评

暂无评论

Potential Transmission Choice for Internet of Things(IoT):Wireless and Batteryless Communications and Open Problems

引用

China Communications 2021年第2期18卷 241-249页

作者： Zhan Xu Guanjie Hu Minzheng Jia Lan Dong School of Information and Communication Engineering and with Key Laboratory of Modern Measurement&Control Technology Ministry of EducationBeijing Information Science&Technology UniversityChina Beijing Key Lab of Transportation Data Analysis and Mining School of Computer and Information TechnologyBeijing Jiaotong UniversityBeijingChina Department of Information Engineering Beijing Polytechnic CollegeBeijingChina

The 5th generation mobile communications aims at connecting everything and future Internet of Things(IoT)will get everything smartly *** realize it,there exist many *** key challenge is the battery problem for small devices,such as sensors or *** backscatter,also referred to as or battery-free backscatter,is a new potential technology to address this *** early and typical type of batteryless backscatter is ambient ***,batteryless backscatter utilizes environmental wireless signals to enable battery-free devices to communicate with each *** devices first harvest energy from ambient wireless signals and then backscatter these signals so as to transmit their own *** paper reviews the current studies about batteryless backscatter,including various backscatter schemes and theoretical works,and then introduces open problems for future research.

关键词： batteryless backscatter battery-free channel state information(CSI) channel estimation multiple antennas signal detection symbiotic communication

来源：评论

学校读者我要写书评

暂无评论

Non-generative Generalized Zero-shot Learning via Task-correlated Disentanglement and Controllable Samples Synthesis

arXiv

引用

arXiv 2022年

作者： Feng, Yaogong Huang, Xiaowen Yang, Pengbo Yu, Jian Sang, Jitao School of Computer and Information Technology Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University China

Synthesizing pseudo samples is currently the most effective way to solve the Generalized Zero-Shot Learning (GZSL) problem. Most models achieve competitive performance but still suffer from two problems: (1) Feature confounding, the overall representations confound task-correlated and task-independent features, and existing models disentangle them in a generative way, but they are unreasonable to synthesize reliable pseudo samples with limited samples;(2) Distribution uncertainty, that massive data is needed when existing models synthesize samples from the uncertain distribution, which causes poor performance in limited samples of seen classes. In this paper, we propose a non-generative model to address these problems correspondingly in two modules: (1) Task-correlated feature disentanglement, to exclude the task-correlated features from task-independent ones by adversarial learning of domain adaption towards reasonable synthesis;(2) Controllable pseudo sample synthesis, to synthesize edge-pseudo and center-pseudo samples with certain characteristics towards more diversity generated and intuitive transfer. In addation, to describe the new scene that is the limit seen class samples in the training process, we further formulate a new ZSL task named the'Few-shot Seen class and Zero-shot Unseen class learning' (FSZU). Extensive experiments on four benchmarks verify that the proposed method is competitive in the GZSL and the FSZU tasks. Copyright © 2022, The Authors. All rights reserved.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

A LLM-based Controllable, Scalable, Human-Involved User Simulator Framework for Conversational Recommender Systems

arXiv

引用

arXiv 2024年

作者： Zhu, Lixi Huang, Xiaowen Sang, Jitao School of Computer Science and Technology Beijing Jiaotong University Beijing China School of Computer Science and Technology Beijing Jiaotong University Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University Key Laboratory of Big Data & Artificial Intelligence in Transportation Beijing Jiaotong University Ministry of Education Beijing China

Conversational Recommender System (CRS) leverages real-time feedback from users to dynamically model their preferences, thereby enhancing the system’s ability to provide personalized recommendations and improving the overall user experience. CRS has demonstrated significant promise, prompting researchers to concentrate their efforts on developing user simulators that are both more realistic and trustworthy. The emergence of Large Language Models (LLMs) has marked the onset of a new epoch in computational capabilities, exhibiting human-level intelligence in various tasks. Research efforts have been made to utilize LLMs for building user simulators to evaluate the performance of CRS. Although these efforts showcase innovation, they are accompanied by certain limitations. In this work, we introduce a Controllable, Scalable, and Human-Involved (CSHI) simulator framework that manages the behavior of user simulators across various stages via a plugin manager. CSHI customizes the simulation of user behavior and interactions to provide a more lifelike and convincing user interaction experience. Through experiments and case studies in two conversational recommendation scenarios, we show that our framework can adapt to a variety of conversational recommendation settings and effectively simulate users’ personalized preferences. Consequently, our simulator is able to generate feedback that closely mirrors that of real users. This facilitates a reliable assessment of existing CRS studies and promotes the creation of high-quality conversational recommendation datasets. © 2024, CC BY-NC-ND.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Towards Alleviating the Object Bias in Prompt Tuning-based Factual Knowledge Extraction

arXiv

引用

arXiv 2023年

作者： Wang, Yuhang Lu, Dongyuan Kong, Chao Sang, Jitao Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University Beijing China School of Information Technology and Management University of International Business and Economics Beijing China

Many works employed prompt tuning methods to automatically optimize prompt queries and extract the factual knowledge stored in Pretrained Language Models. In this paper, we observe that the optimized prompts, including discrete prompts and continuous prompts, exhibit undesirable object bias. To handle this problem, we propose a novel prompt tuning method called MeCoD1. consisting of three modules: Prompt Encoder, Object Equalization and Biased Object Obstruction. Experimental results show that MeCoD can significantly reduce the object bias and at the same time improve accuracy of factual knowledge extraction. Copyright © 2023, The Authors. All rights reserved.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

Cross-Entropy Loss for Recommending Efficient Fold-Over Technique

引用

Journal of Systems Science & Complexity 2021年第1期34卷 402-439页

作者： WENG Lin-Chen ELSAWAH A M FANG Kai-Tai Division of Science and Technology Beijing Normal University-Hong Kong Baptist University United International CollegeZhuhai 519085China Department of Mathematics Faculty of ScienceZagazig UniversityZagazig 44519Egypt The Key Lab of Random Complex Structures and Data Analysis Chinese Academy of SciencesBeijing 100190China

Due to the limited resources and budgets in many real-life projects, it is unaffordable to use full factorial experimental designs and thus fractional factorial(FF) designs are used instead. The aliasing of factorial effects is the price we pay for using FF designs and thus some significant effects cannot be estimated. Therefore, some additional observations(runs) are needed to break the linages among the factorial effects. Folding over the initial FF designs is one of the significant approaches for selecting the additional runs. This paper gives an in-depth look at fold-over techniques via the following four significant contributions. The first contribution is on discussing the adjusted switching levels foldover technique to overcome the limitation of the classical one. The second contribution is on presenting a comparison study among the widely used fold-over techniques to help experimenters to recommend a suitable fold-over technique for their experiments by answering the following two fundamental questions:Do these techniques dramatically lessen the confounding of the initial designs, and do the resulting combined designs(combining initial design with its fold-over) via these techniques have considerable difference from the optimality point of view considering the markedly different searching domains in each technique? The optimality criteria are the aberration, confounding, Hamming distance and uniformity. Many of these criteria are given in sequences(patterns) form, which are inconvenient and costly to represent and compare, especially when the designs have many factors. The third innovation is on developing a new criterion(dictionary cross-entropy loss) to simplify the existing criteria fromsequence to scalar. The new criterion leads to a more straightforward and easy comparison study. The final contribution is on establishing a general framework for the connections between initial designs and combined designs based on any fold-over technique.

关键词： Aberration confounding cross-entropy dictionary ordering fold-over Hamming distance loss function uniformity

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：