检索结果-内蒙古大学图书馆

Hybrid Data-Free Knowledge Distillation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Tang, Jialiang Chen, Shuo Gong, Chen School of Computer Science and Engineering Nanjing University of Science and Technology China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education China Jiangsu Key Laboratory of Image and Video Understanding for Social Security China Center for Advanced Intelligence Project RIKEN Japan Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University China

Data-free knowledge distillation aims to learn a compact student network from a pre-trained large teacher network without using the original training data of the teacher network. Existing collection-based and generation-based methods train student networks by collecting massive real examples and generating synthetic examples, respectively. However, they inevitably become weak in practical scenarios due to the difficulties in gathering or emulating sufficient real-world data. To solve this problem, we propose a novel method called Hybrid Data-Free Distillation (HiDFD), which leverages only a small amount of collected data as well as generates sufficient examples for training student networks. Our HiDFD comprises two primary modules, i.e., the teacher-guided generation and student distillation. The teacher-guided generation module guides a Generative Adversarial Network (GAN) by the teacher network to produce high-quality synthetic examples from very few real-world collected examples. Specifically, we design a feature integration mechanism to prevent the GAN from overfitting and facilitate the reliable representation learning from the teacher network. Meanwhile, we drive a category frequency smoothing technique via the teacher network to balance the generative training of each category. In the student distillation module, we explore a data inflation strategy to properly utilize a blend of real and synthetic data to train the student network via a classifier-sharing-based feature alignment technique. Intensive experiments across multiple benchmarks demonstrate that our HiDFD can achieve state-of-the-art performance using 120 times less collected data than existing methods. Code is available at https://***/tangjialiang97/HiDFD. Copyright © 2024, The Authors. All rights reserved.

关键词： Students

New intra prediction mode for interlaced video in AVS

学校读者我要写书评

暂无评论

Gaojishu Tongxin/Chinese High Technology Letters 2007年第3期17卷 221-225页

作者： Chen, Liang Tan, Yihua Tian, Jinwen State Key Laboratory of Education Ministry for Image Processing and Intelligent Control Institute of Pattern Recognition and Artificial Intelligence Huazhong University of Science and Technology Wuhan 430074 China

The technologies of intra prediction and MBAFF were introduced, and a new intra prediction mode based on the characteristics of spatial distribution in interlaced video was proposed. The spatial correlation of five luma intra prediction modes in AVS-P2 and the new mode were analyzed. From the analysis result, it can be concluded that the new mode can exploit the spatial correlation better and predict the samples more precisely than the existed ones. The experimental results showed that the average gain in peak signal to noise ratio was above 0.12dB and the average reduction in bit-rate was above 1.77%, so the proposed mode is an effective prediction mode for improvement of coding performance.

关键词： AVS Interlaced video Intra prediction Video coding

Study on the Method of Image Restoration in the Environment of Dust

学校读者我要写书评

暂无评论

Study on the Method of Image Restoration in the Environment ...

International Conference on Wireless Communications, Networking and Mobile Computing (WiCom)

作者： Yuanyu Wang Yuanzong Li Tianxu Zhang Department of Computer Fundamental Education Taiyuan University Of Technology Taiyuan China College of Mechanical Engineering Taiyuan University Of Technology Taiyuan China Institute of Pattern Recognition and A.I State Key Laboratory of Image Processing and Intelligent Control HuaZhong University of Science and Technology Wuhan China

In order to solve the problem of image degradation caused by dust environments, an image degradation model considering multiple scattering factors caused by dust was first established using the first-order multiple scattering method. Then, a dark channel prior principle was applied to present an image restoration algorithm based on the image degradation model. Finally, GA optimization algorithm was applied to optimize the atmospheric light and the exposure parameters. This optimization algorithm was established according to the criterion of the image evaluation based on kirsch operator with automatic threshold. By using the method an optimistic result of image restoration was obtained. The experimental results have shown that the method not only enhanced luminance and contrast, but also discovered more detail edges information. The method provided a foundation for target recognition in the dust environments.

关键词： Image edge detection Image restoration Scattering Degradation Gallium Atmospheric modeling Genetic algorithms

Incremental transformer with deliberation decoder for document grounded conversations

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Li, Zekang Niu, Cheng Meng, Fandong Feng, Yang Li, Qian Zhou, Jie Dian Group School of Electronic Information and Communications Huazhong University of Science and Technology Pattern Recognition Center WeChat AI Tencent Inc China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences School of Computer Science and Engineering Northeastern University China

Document Grounded Conversations is a task to generate dialogue responses when chatting about the content of a given document. Obviously, document knowledge plays a critical role in Document Grounded Conversations, while existing dialogue models do not exploit this kind of knowledge effectively enough. In this paper, we propose a novel Transformerbased architecture for multi-turn document grounded conversations. In particular, we devise an Incremental Transformer to encode multi-turn utterances along with knowledge in related documents. Motivated by the human cognitive process, we design a two-pass decoder (Deliberation Decoder) to improve context coherence and knowledge correctness. Our empirical study on a real-world Document Grounded Dataset proves that responses generated by our model significantly outperform competitive baselines on both context coherence and knowledge relevance. Copyright © 2019, The Authors. All rights reserved.

关键词： Decoding

Fast signal recovery from saturated measurements by linear loss and nonconvex penalties

学校读者我要写书评

暂无评论

arXiv 2018年

作者： He, Fan Huang, Xiaolin Liu, Yipeng Yan, Ming Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University The MOE Key Laboratory of System Control and Information Processing Shanghai200240 China School of Information and Communication Engineering University of Electronic Science and Technology of China Chengdu611731 China The Department of Computational Mathematics Science and Engineering Michigan State University MI United States

Sign information is the key to overcoming the inevitable saturation error in compressive sensing systems, which causes information loss and results in bias. For sparse signal recovery from saturation, we propose to use a linear loss to improve the effectiveness from existing methods that utilize hard constraints/hinge loss for sign consistency. Due to the use of linear loss, an analytical solution in the update progress is obtained, and some nonconvex penalties are applicable, e.g., the minimax concave penalty, the 0 norm, and the sorted 1 norm. Theoretical analysis reveals that the estimation error can still be bounded. Generally, with linear loss and nonconvex penalties, the recovery performance is significantly improved, and the computational time is largely saved, which is verified by the numerical experiments. Copyright © 2018, The Authors. All rights reserved.

关键词： Compressed sensing

Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based Agents

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Wenkai Bi, Xiaohan Lin, Yankai Chen, Sishuo Zhou, Jie Sun, Xu Gaoling School of Artificial Intelligence Renmin University of China Beijing China Center for Data Science Peking University China Pattern Recognition Center WeChat AI Tencent Inc. China National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University China

Driven by the rapid development of Large Language Models (LLMs), LLM-based agents have been developed to handle various real-world applications, including finance, healthcare, and shopping, etc. It is crucial to ensure the reliability and security of LLM-based agents during applications. However, the safety issues of LLM-based agents are currently under-explored. In this work, we take the first step to investigate one of the typical safety threats, backdoor attack, to LLM-based agents. We first formulate a general framework of agent backdoor attacks, then we present a thorough analysis of different forms of agent backdoor attacks. Specifically, compared with traditional backdoor attacks on LLMs that are only able to manipulate the user inputs and model outputs, agent backdoor attacks exhibit more diverse and covert forms: (1) From the perspective of the final attacking outcomes, the agent backdoor attacker can not only choose to manipulate the final output distribution, but also introduce the malicious behavior in an intermediate reasoning step only, while keeping the final output correct. (2) Furthermore, the former category can be divided into two subcategories based on trigger locations, in which the backdoor trigger can either be hidden in the user query or appear in an intermediate observation returned by the external environment. We implement the above variations of agent backdoor attacks on two typical agent tasks including web shopping and tool utilization. Extensive experiments show that LLM-based agents suffer severely from backdoor attacks and such backdoor vulnerability cannot be easily mitigated by current textual backdoor defense algorithms. This indicates an urgent need for further research on the development of targeted defenses against backdoor attacks on LLM-based agents.3 Warning: This paper may contain biased content. © 2024, CC BY.

关键词：

Lower bounds on some van der waerden numbers based on quadratic residues

学校读者我要写书评

暂无评论

Journal of Combinatorial Mathematics and Combinatorial Computing 2012年 83卷 23-31页

作者： Liang, Meilian Xu, Xiaodong Shao, Zehui Xiu, Baoxin School of Mathematics and Information Science Guangxi University Nanning 530004 China Guangxi Academy of Sciences Nanning 530007 China Key Laboratory of Pattern Recognition and Intelligent Information Processing School of Information Science and Technology Chengdu University Chengdu 610106 China College of Information System and Management National University of Defense Technology Changsha 410073 China

The van der Waerden number W(r, k) is the least integer N such that every r-coloring of {1,2, • • •, N} contains a monochromatic arithmetic progression of length at least k. Rabung gave a method to obtain lower bounds on W(2,k) based on quadratic residues, and performed computations on all primes no greater than 20117. By improving the efficiency of the algorithm of Rabung, we perform the computation for all primes up to 6 x 107, and obtain lower bounds on W(2, k) for k between 11 and 23.

关键词： Computational efficiency

Translatotron-V(ison): An End-to-End Model for In-Image Machine Translation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Lan, Zhibin Niu, Liqiang Meng, Fandong Zhou, Jie Zhang, Min Su, Jinsong School of Informatics Xiamen University China Pattern Recognition Center WeChat AI Tencent Inc China Key Laboratory of Digital Protection and Intelligent Processing of Intangible Cultural Heritage of Fujian and Taiwan Xiamen University Ministry of Culture and Tourism China Institute of Computer Science and Technology Soochow University China

In-image machine translation (IIMT) aims to translate an image containing texts in source language into an image containing translations in target language. In this regard, conventional cascaded methods suffer from issues such as error propagation, massive parameters, and difficulties in deployment and retaining visual characteristics of the input image. Thus, constructing end-to-end models has become an option, which, however, faces two main challenges: 1) the huge modeling burden, as it is required to simultaneously learn alignment across languages and preserve the visual characteristics of the input image;2) the difficulties of directly predicting excessively lengthy pixel sequences. In this paper, we propose Translatotron-V(ision), an end-to-end IIMT model consisting of four modules. In addition to an image encoder, and an image decoder, our model contains a target text decoder and an image tokenizer. Among them, the target text decoder is used to alleviate the language alignment burden, and the image tokenizer converts long sequences of pixels into shorter sequences of visual tokens, preventing the model from focusing on low-level visual features. Besides, we present a two-stage training framework for our model to assist the model in learning alignment across modalities and languages. Finally, we propose a location-aware evaluation metric called Structure-BLEU to assess the translation quality of the generated images. Experimental results demonstrate that our model achieves competitive performance compared to cascaded models with only 70.9% of parameters, and significantly outperforms the pixel-level end-to-end IIMT model. Copyright © 2024, The Authors. All rights reserved.

关键词： Pixels

Learning data-adaptive nonparametric kernels

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Fanghui Huang, Xiaolin Gong, Chen Yang, Jie Li, Li Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai200240 China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China Department of Automation Tsinghua University

Kernel methods have been extensively used in a variety of machine learning tasks such as classification, clustering, and dimensionality reduction. For complicated practical tasks, the traditional kernels, e.g., Gaussian kernel and sigmoid kernel, or their combinations are often not sufficiently flexible to fit the data. In this paper, we present a Data-Adaptive Nonparametric Kernel (DANK) learning framework in a data-driven manner. To be specific, in model formulation, we impose an adaptive matrix on the kernel/Gram matrix in an entry-wise strategy. Since we do not specify the formulation of the adaptive matrix, each entry in the adaptive matrix can be directly and flexibly learned from the data. Therefore, the solution space of the learned kernel is largely expanded, which makes our DANK model flexible to capture the data with different local statistical properties. Specifically, the proposed kernel learning framework can be seamlessly embedded to support vector machines (SVM) and support vector regression (SVR), which has the capability of enlarging the margin between classes and reducing the model generalization error. Theoretically, we demonstrate that the objective function of our DANK model embedded in SVM/SVR is gradient-Lipschitz continuous. Thereby, the training process for kernel and parameter learning in SVM/SVR can be efficiently optimized in a unified framework. Further, to address the scalability issue in nonparametric kernel learning framework, we decompose the entire optimization problem in DANK into several smaller easy-to-solve problems, so that our DANK model can be efficiently approximated by this partition. The effectiveness of this approximation is demonstrated by both empirical studies and theoretical guarantees. Experimentally, the proposed DANK model embedded in SVM/SVR achieves encouraging performance on various classification and regression benchmark datasets when compared with other representative kernel learning based algorithms. Copyrig

关键词： Matrix algebra