检索结果-内蒙古大学图书馆

DFMC:Feature-Driven data-Free Knowledge Distillation

IEEE Transactions on Circuits and Systems for Video technology 2025年

作者： Zhang, Zherui Xu, Rongtao Wang, Changwei Xu, Wenhao Chen, Shunpeng Xu, Shibiao Xu, Guangyuan Guo, Li Beijing University of Posts and Telecommunications School of Artificial Intelligence Beijing100876 China Chinese Academy of Sciences State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Beijing100190 China Qilu University of Technology Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center Jinan250316 China Shandong Fundamental Research Center for Computer Science Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Jinan250014 China

data-Free Knowledge Distillation (DFKD) enables knowledge transfer from teacher networks without access to the real dataset. However, generator-based DFKD methods often suffer from insufficient diversity or low-confidence in synthetic images, negatively impacting student network performance. This paper introduces DFMC, a generative feature-driven framework to mitigate the inherent limitations of DFKD. We propose exploiting semantic description between generative feature domains to guide augmentation strategies, avoiding random abstract inputs caused by inconsistent semantic quality. Then, by applying noise to the generative features, we produce contrastive learning pairs indirectly, limiting the sampling range of the feature domain to encourage the student network to learn domain-invariant features. Finally, we guide the student network to deeply mimic the teacher’s layer-wise implicit classification behavior for the augmented synthetic images. Extensive experiments across various datasets and downstream tasks demonstrate the effectiveness of DFMC, achieving significant improvements while preventing student networks from overfitting to semantic ambiguous images. © 1991-2012 IEEE.

关键词： Knowledge transfer

来源：评论

学校读者我要写书评

暂无评论

Uncertainty calibration for ensemble-based debiasing methods

arXiv

引用

arXiv 2021年

作者： Xiong, Ruibin Chen, Yimeng Pang, Liang Cheng, Xueqi Ma, Zhiming Lan, Yanyan CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Baidu Inc Academy of Mathematics and Systems Science Chinese Academy of Sciences Data Intelligence System Research Center Institute of Computing Technology Chinese Academy of Sciences Institute for AI Industry Research Tsinghua University

Ensemble-based debiasing methods have been shown effective in mitigating the reliance of classifiers on specific dataset bias, by exploiting the output of a bias-only model to adjust the learning target. In this paper, we focus on the bias-only model in these ensemble-based methods, which plays an important role but has not gained much attention in the existing literature. Theoretically, we prove that the debiasing performance can be damaged by inaccurate uncertainty estimations of the bias-only model. Empirically, we show that existing bias-only models fall short in producing accurate uncertainty estimations. Motivated by these findings, we propose to conduct calibration on the bias-only model, thus achieving a three-stage ensemble-based debiasing framework, including bias modeling, model calibrating, and debiasing. Experimental results on NLI and fact verification tasks show that our proposed three-stage debiasing framework consistently outperforms the traditional two-stage one in out-of-distribution accuracy. Copyright © 2021, The Authors. All rights reserved.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Link Prediction on n-ary relational data based on relatedness evaluation

arXiv

引用

arXiv 2021年

作者： Guan, Saiping Jin, Xiaolong Guo, Jiafeng Wang, Yuanzhuo Cheng, Xueqi School of Computer Science and Technology University of Chinese Academy of Sciences Beijing100049 China CAS Key Lab of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences Beijing100864 China

With the overwhelming popularity of Knowledge Graphs (KGs), researchers have poured attention to link prediction to fill in missing facts for a long time. However, they mainly focus on link prediction on binary relational data, where facts are usually represented as triples in the form of (head entity, relation, tail entity). In practice, n-ary relational facts are also ubiquitous. When encountering such facts, existing studies usually decompose them into triples by introducing a multitude of auxiliary virtual entities and additional triples. These conversions result in the complexity of carrying out link prediction on n-ary relational data. It has even proven that they may cause loss of structure information. To overcome these problems, in this paper, we represent each n-ary relational fact as a set of its role and role-value pairs. We then propose a method called NaLP to conduct link prediction on n-ary relational data, which explicitly models the relatedness of all the role and role-value pairs in an n-ary relational fact. We further extend NaLP by introducing type constraints of roles and role-values without any external type-specific supervision, and proposing a more reasonable negative sampling mechanism. Experimental results validate the effectiveness and merits of the proposed methods. Copyright © 2021, The Authors. All rights reserved.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

A Constraints Fusion-induced Symmetric Nonnegative Matrix Factorization Approach for Community Detection

arXiv

引用

arXiv 2023年

作者： Liu, Zhigang Luo, Xin College of Computer Science and Technology Chongqing University of Posts and Telecommunications Chongqing400065 China Chongqing Key Laboratory of Big Data and Intelligent Computing Chongqing Institute of Green and Intelligent Technology Chinese Academy of Sciences Chongqing400714 China College of Computer and Information Science Southwest University Chongqing400715 China

Community is a fundamental and critical characteristic of an undirected social network, making community detection be a vital yet thorny issue in network representation learning. A symmetric and non-negative matrix factorization (SNMF) model is frequently adopted to address this issue owing to its great interpretability and scalability. However, it adopts a single latent factor matrix to represent an undirected network for precisely representing its symmetry, which leads to loss of representation learning ability due to the reduced latent space. Motivated by this discovery, this paper proposes a novel Constraints Fusion-induced Symmetric Nonnegative Matrix Factorization (CFS) model that adopts three-fold ideas: a) Representing a target undirected network with multiple latent factor matrices, thus preserving its representation learning capacity;b) Incorporating a symmetry-regularizer that preserves the symmetry of the learnt low-rank approximation to the adjacency matrix into the loss function, thus making the resultant detector well-aware of the target network's symmetry;and c) Introducing a graph-regularizer that preserves local invariance of the network's intrinsic geometry, thus making the achieved detector well-aware of community structure within the target network. Extensively empirical studies on eight real-world social networks from industrial applications demonstrate that the proposed CFS model significantly outperforms state-of-the-art models in achieving highly-accurate community detection results. Copyright © 2023, The Authors. All rights reserved.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

Scaling up Dynamic Edge Partition Models via Stochastic Gradient MCMC

arXiv

引用

arXiv 2024年

作者： Yang, Sikun Koeppl, Heinz School of Computing and Information Technology Great Bay University Dongguan523000 China Great Bay Institute for Advanced Study Great Bay University Dongguan Key Laboratory for Data Science and Intelligent Medicine Great Bay University China Technische Universität Darmstadt Darmstadt Germany

The edge partition model (EPM) is a generative model for extracting an overlapping community structure from static graph-structured data. In the EPM, the gamma process (GaP) prior is adopted to infer the appropriate number of latent communities, and each vertex is endowed with a gamma distributed positive memberships vector. Despite having many attractive properties, inference in the EPM is typically performed using Markov chain Monte Carlo (MCMC) methods that prevent it from being applied to massive network data. In this paper, we generalize the EPM to account for dynamic enviroment by representing each vertex with a positive memberships vector constructed using Dirichlet prior specification, and capturing the time-evolving behaviour of vertices via a Dirichlet Markov chain construction. A simple-to-implement Gibbs sampler is proposed to perform posterior computation using Negative-Binomial augmentation technique. For large network data, we propose a stochastic gradient Markov chain Monte Carlo (SG-MCMC) algorithm for scalable inference in the proposed model. The experimental results show that the novel methods achieve competitive performance in terms of link prediction, while being much faster. © 2024, CC BY.

关键词： Monte Carlo methods

来源：评论

学校读者我要写书评

暂无评论

LMVD: A Large-Scale Multimodal Vlog dataset for Depression Detection in the Wild

arXiv

引用

arXiv 2024年

作者： He, Lang Chen, Kai Zhao, Junnan Wang, Yimeng Pei, Ercheng Chen, Haifeng Jiang, Jiewei Zhang, Shiqing Zhang, Jie Wang, Zhongmin He, Tao Tiwari, Prayag The School of Computer Science and Technology Xi’an University of Posts and Telecommunications Shaanxi Xi’an710121 China The Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Shaanxi Xi’an710121 China The Xi’an Key Laboratory of Big Data and Intelligent Computing Shaanxi Xi’an710121 China The School of Electronic Engineering Xi’an University of Posts and Telecommunications Xi’an710121 China The College of Electronic Information and Artificial Intelligence Shaanxi University of Science and Technology Xi’an710016 China The Institute of Intelligent Information Processing Taizhou University Zhejiang Taizhou318000 China The School of Education Shenzhen University Shenzhen518060 China The School of Information Technology Halmstad University Halmstad Sweden

Depression can significantly impact many aspects of an individual’s life, including their personal and social functioning, academic and work performance, and overall quality of life. Many researchers within the field of affective computing are adopting deep learning technology to explore potential patterns related to the detection of depression. However, because of subjects’ privacy protection concerns, that data in this area is still scarce, presenting a challenge for the deep discriminative models used in detecting depression. To navigate these obstacles, a large-scale multimodal vlog dataset (LMVD), for depression recognition in the wild is built. In LMVD, which has 1823 samples with 214 hours of the 1475 participants captured from four multimedia platforms (Sina Weibo, Bilibili, Tiktok, and YouTube). A novel architecture termed MDDformer to learn the non-verbal behaviors of individuals is proposed. Extensive validations are performed on the LMVD dataset, demonstrating superior performance for depression detection. We anticipate that the LMVD will contribute a valuable function to the depression detection community. The data and code will released at the link: https://***/helang818/LMVD/. © 2024, CC BY-NC-ND.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Optimizing FPGA-Based DCN Accelerator with On-Chip dataflow Reordering and Serial-Parallel computing Array

Optimizing FPGA-Based DCN Accelerator with On-Chip Dataflow ...

引用

High Performance Big data and Intelligent Systems (HPBD&IS)

作者： Ming Zhang Jian Xu Jinzhong He Hong Qin AnnLab Institute of Semiconductors Chinese Academy of Sciences Beijing China Beijing Key Laboratory of Semiconductor Neural Network Intelligent Sensing and Computing Technology Beijing China College of Materials Science and Opto-Electronic Technology & School of Integrated Circuits University of Chinese Academy of Sciences Beijing China

Deep convolutional neural networks (DCNs) have recently experienced rapid development in the direction of lightweight and edge deployment. However, accelerators for DCNs face challenges in balancing computational and data bandwidth, leading to inefficient computation and high hardware costs. Additionally, different network structures make it challenging to design and reconfigure accelerators flexibly. To address these issues, this paper proposes a parallel-serial channel accelerator system, which resolves the low utilization of multipliers caused by small channels and inadequate bandwidth of fully connected layers. The results demonstrate that the proposed accelerator in this study maintains high computational performance and efficiency on typical DCNs. When implemented on Xilinx VCU128 at 200 MHz, the peak computational performance reaches 204.5 GOPS, with an efficiency of 0.37 GOPS/DSP and a maximum utilization rate of computing array up to 99.63%, surpassing previous works.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Generalization Boosted Adapter for Open-Vocabulary Segmentation

arXiv

引用

arXiv 2024年

作者： Xu, Wenhao Wang, Changwei Feng, Xuxiang Xu, Rongtao Huang, Longzhao Zhang, Zherui Guo, Li Xu, Shibiao School of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing100876 China The Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center National Supercomputer Center in Jinan Qilu University of Technology Shandong Academy of Sciences Jinan250013 China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Jinan China The Aerospace Information Research Institute Chinese Academy of Sciences China The State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing China

Vision-language models (VLMs) have demonstrated remarkable open-vocabulary object recognition capabilities, motivating their adaptation for dense prediction tasks like segmentation. However, directly applying VLMs to such tasks remains challenging due to their lack of pixel-level granularity and the limited data available for fine-tuning, leading to overfitting and poor generalization. To address these limitations, we propose Generalization Boosted Adapter (GBA), a novel adapter strategy that enhances the generalization and robustness of VLMs for open-vocabulary segmentation. GBA comprises two core components: (1) a Style Diversification Adapter (SDA) that decouples features into amplitude and phase components, operating solely on the amplitude to enrich the feature space representation while preserving semantic consistency;and (2) a Correlation Constraint Adapter (CCA) that employs cross-attention to establish tighter semantic associations between text categories and target regions, suppressing irrelevant low-frequency "noise" information and avoiding erroneous associations. Through the synergistic effect of the shallow SDA and the deep CCA, GBA effectively alleviates overfitting issues and enhances the semantic relevance of feature representations. As a simple, efficient, and plug-and-play component, GBA can be flexibly integrated into various CLIP-based methods, demonstrating broad applicability and achieving state-of-the-art performance on multiple open-vocabulary segmentation benchmarks. © 2024, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A Novel Cloud-Native and Multi-Platform Parallelized SBAS-INSAR Algorithm

A Novel Cloud-Native and Multi-Platform Parallelized SBAS-IN...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Peichen Yu Chao Wang Yixian Tang Weikang Zhang Haihang You Shaoyang Guan Lichuan Zou Hong Zhang Key Laboratory of Digital Earth Science Aerospace Information Research Institute Chinese Academy of Sciences Beijing China International Research Center of Big Data for Sustainable Development Goals Beijing China University of Chinese Academy of Sciences Beijing China Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (数字)9798350360325

ISBN: (纸本)9798350360332

Facing massive Synthetic Aperture Radar (SAR) data, the challenge of achieving rapid and efficient data processing has garnered attention. Currently, most time-series InSAR processing solutions are designed to operate on a single computing platform, leading to drawbacks such as inflexible deployment and slow data transmission. Leveraging the open-source Kubernetes platform, we have employed cloud-native technology to deploy a cross-platform, multi-level parallelization algorithm across multiple nodes. The algorithm is designed based on modular concept, and its modules can successfully achieve cross-platform deployment by relying on both the network File System (NFS) protocol and the Common Internet File System (CFIS) protocol. This overcomes the limitations of the previous data file storage system, which solely depended on the NFS protocol, leading to deployment on a single operating system. Deploying the algorithm on different operating systems, our results show that the speeds of Linux platform's algorithm parallel modules were improved 16% and 30%, respectively. The successful cross-platform operation of the algorithm enables the data processing workflow to assimilate the strengths of different operating platforms, enhancing data visualization capabilities and gaining support from diverse platform resources. This introduces a new approach for large-scale time-series InSAR processing.

关键词： Cloud computing Protocols File systems Operating systems Linux Geoscience and remote sensing data visualization

来源：评论

学校读者我要写书评

暂无评论

Adaptive information seeking for open-domain question answering

arXiv

引用

arXiv 2021年

作者： Zhu, Yunchang Pang, Liang Lan, Yanyan Shen, Huawei Cheng, Xueqi Data Intelligence System Research Center CAS Key Lab of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Institute for AI Industry Research Tsinghua University

Information seeking is an essential step for open-domain question answering to efficiently gather evidence from a large corpus. Recently, iterative approaches have been proven to be effective for complex questions, by recursively retrieving new evidence at each step. However, almost all existing iterative approaches use predefined strategies, either applying the same retrieval function multiple times or fixing the order of different retrieval functions, which cannot fulfill the diverse requirements of various questions. In this paper, we propose a novel adaptive information-seeking strategy for open-domain question answering, namely AISO. Specifically, the whole retrieval and answer process is modeled as a partially observed Markov decision process, where three types of retrieval operations (e.g., BM25, DPR, and hyperlink) and one answer operation are defined as actions. According to the learned policy, AISO could adaptively select a proper retrieval action to seek the missing evidence at each step, based on the collected evidence and the reformulated query, or directly output the answer when the evidence set is sufficient for the question. Experiments on SQuAD Open and HotpotQA fullwiki, which serve as single-hop and multi-hop open-domain QA benchmarks, show that AISO outperforms all baseline methods with predefined strategies in terms of both retrieval and answer evaluations. © 2021, CC BY.

关键词： Hypertext systems

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

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

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：