检索结果-内蒙古大学图书馆

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

来源：评论

学校读者我要写书评

暂无评论

Hybrid Data-Free Knowledge Distillation

arXiv

引用

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

来源：评论

学校读者我要写书评

暂无评论

Cross-Modal De-Deviation for Enhancing Few-Shot Classification

SSRN

引用

SSRN 2023年

作者： Pan, Mei-Hong Shen, Hong-Bin School of Electronic Information and Electrical Engineering Shanghai Jiaotong University China Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai200240 China

Few-shot learning poses a critical challenge due to the deviation problem caused by the scarcity of available samples. In this work, we aim to address deviations in both feature representations and prototypes. To achieve this, we propose a cross-modal de-deviation framework that leverages class semantic information to provide robust prior knowledge for the samples. This framework begins with a visual-to-semantic autoencoder trained on the labeled samples to predict semantic features for the unlabeled samples. Then, we devise a binary linear programming model to incorporate the initial prototypes with the cluster centers of the unlabeled samples. To avoid mismatch between the cluster centers and the initial prototypes, we conduct the label assignment process in the semantic space using the class ground truth semantic features as the reference points, with the cluster centers transformed into semantic representations. Moreover, we model a linear classifier with the concatenation of the refined prototypes and the class ground truth semantic features serving as the initial weights. Then we propose a novel optimization strategy based on the alternating least squares (ALS) model. From the ALS model, we can derive two closed-form solutions regarding to the features and weights, facilitating alternative optimization of them. Extensive experiments conducted on three standard benchmarks demonstrate the competitive advantages of our CMDD method over the state-of-the-art few-shot classification methods, confirming its effectiveness in reducing deviation. © 2023, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A Semantic Segmentation Method of Buildings in Remote Sensing Image Based on Improved UNet 2

A Semantic Segmentation Method of Buildings in Remote Sensin...

引用

2nd International Conference on Signal Image Processing and Communication, ICSIPC 2022

作者： Li, Zhongyu Liu, Yang Kuang, Yin Wang, Huajun Liu, Cheng College of Computer Science Chengdu Normal University Chengdu611130 China College of Geophysics Chengdu University of Technology Chengdu610059 China Key Laboratory of Pattern Recognition and Intelligent Information Processing of Sichuan Chengdu University Chengdu610106 China Artificial Intelligence Key Laboratory of Sichuan Province Zigong643000 China Key Laboratory of interior Layout optimization and Security Institutions of Higher Education of Sichuan Province Chengdu Normal University Sichuan Chengdu611130 China College of Movie and Media Sichuan Normal University Chengdu610066 China

ISBN: (纸本)9781510657694

Aiming at the problem of model instability and overfitting of deep neural networks with the deepening of the number of network layers, the current mainstream method is to use batch normalization (BN) to alleviate them. However, since the BN method is more sensitive to batch size when the batch size is small, the model performance will be poor. For a relatively large model, due to the limitation of video memory, the batch size cannot take a large value, limiting the model's performance. Because of the dependence of BN on batch size, this paper adopts group normalization (GN) instead of batch normalization (BN) in the UNet network to alleviate the impact of the model on batch size. Then experiments are carried out on the WHUBuilding dataset. The experimental results show that the improved model (UNet-GN) improves the mean intersection over union (MIoU) and mean pixel accuracy (MPA) by 10.66% and 1.65% respectively compared with the original model (UNet-BN). © 2022 SPIE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Unifying Gradients to Improve Real-World Robustness for Deep Networks

arXiv

引用

arXiv 2022年

作者： Wu, Yingwen Chen, Sizhe Fang, Kun Huang, Xiaolin Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University 800 Dongchuan Road Shanghai China Institute of Image Processing and Pattern Recognition The MOE Key Laboratory of System Control and Information Processing Shanghai Jiao Tong University 800 Dongchuan Road Shanghai China

The wide application of deep neural networks (DNNs) demands an increasing amount of attention to their real-world robustness, i.e., whether a DNN resists black-box adversarial attacks, among which score-based query attacks (SQAs) are most threatening since they can effectively hurt a victim network with the only access to model outputs. Defending against SQAs requires a slight but artful variation of outputs due to the service purpose for users, who share the same output information with SQAs. In this paper, we propose a real-world defense by Unifying Gradients (UniG) of different data so that SQAs could only probe a much weaker attack direction that is similar for different samples. Since such universal attack perturbations have been validated as less aggressive than the input-specific perturbations, UniG protects real-world DNNs by indicating attackers a twisted and less informative attack direction. We implement UniG efficiently by a Hadamard product module which is plug-and-play. According to extensive experiments on 5 SQAs, 2 adaptive attacks and 7 defense baselines, UniG significantly improves real-world robustness without hurting clean accuracy on CIFAR10 and ImageNet. For instance, UniG maintains a model of 77.80% accuracy under 2500-query Square attack while the state-of-the-art adversarially-trained model only has 67.34% on CIFAR10. Simultaneously, UniG outperforms all compared baselines in terms of clean accuracy and achieves the smallest modification of the model output. The code is released at https://***/snowien/UniG-pytorch. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Geometric Numerical Integral Method in Compact Lie Group

Geometric Numerical Integral Method in Compact Lie Group

引用

6th International Conference on intelligent Computing and Signal Processing (ICSP)

作者： Chao Liu Shengyi Yang Key Laboratory of Pattern Recognition and Intelligent System of Guizhou Province Guizhou Minzu University Gui Yang China

ISBN: (纸本)9781665447300

Three dimensions special orthogonal group SO (3) is widely used to describe the rotation kinematics of the rigid body without local coordinates, which can avoid rotation singularity and unwinding in traditional methods. Propagating the rotation kinematics in SO (3) with a specific geometric integration method is not only to obtained numerical results with improved qualitative behavior, but also provided more accurate long-time integration results. While many studies have focused on geometric integration algorithms to preserve the geometric structure, this work has the additional objective of studying result accuracy. Integral curves on SO (3) obtained using the third-order Crouch-Grossman Lie group method are compared with numerical results using the third-order RKMK algorithms, the exponential coordinates method, and the simple projection method. Results show that the use of the Crouch-Grossman Lie group method better preserves the geometric structure of SO (3) for the larger time steps considered. It is also found that the third-order Crouch-Grossman algorithm is more accurate than the RKMK except for the smallest time step used.

关键词： Heuristic algorithms Signal processing algorithms Kinematics Signal processing

来源：评论

学校读者我要写书评

暂无评论

Sn-Cd Modified Ti 2Co 2 MXene for Enhanced Harmful Gas Adsorption: A DFT Study

SSRN

引用

SSRN 2024年

作者： Yu, Wenzhuo Shen, Tao Yuan, Hang Feng, Yue Wu, Haodong Liu, Xin Liu, Chi Heilongjiang Provincial Key Laboratory of Pattern Recognition and Information Perception Harbin University of Science and Technology Harbin150080 China Heilongjiang Provincial Key Laboratory of Quantum Manipulation and Control Harbin University of Science and Technology Harbin150080 China Key Laboratory of Engineering Dielectrics and its Application Ministry of Education Harbin University of Science and Technology Harbin150080 China College of Science Heilongjiang Province Key Laboratory of Laser Spectroscopy Technology and Application Harbin University of Science and Technology Harbin150080 China

In this study, we have investigated the adsorption characteristics of Ti2CO2 monolayer based on density functional theory for the harmful gases (NH3, H2S, CH4, NO2, SO2). The optimum adsorption structure of Ti2CO2 monolayer for adsorption of different gases were determined, and the adsorption mechanism of gases on the material was investigated. NH3 exhibits the highest absolute adsorption energy (0.429 eV) among the gases (NH3, H2S, CH4, NO2, SO2), when physisorbed on Ti2CO2 monolayer. Substitution doping modification of the Ti2CO2 monolayer with the metal element Sn and the transition metal element Cd resulted in alterations to the electronic properties of the materials. The adsorption capacity of the material for NH3, H2S, CH4, NO2, SO2 gases has been improved, which makes the Ti2CO2 monolayer have a broader application prospect in adsorption of toxic and hazardous gases. The substantial enhancement of the adsorption performance of Ti2CO2 monolayers for NO2 (adsorption energy increases to 10.35 times of the original) after doping provides a new idea for doping modification of Ti2CO2 monolayers in order to facilitate the adsorption of different gases. © 2024, The Authors. All rights reserved.

关键词： Monolayers

来源：评论

学校读者我要写书评

暂无评论

Isoform Function Prediction Based on Heterogeneous Graph Attention Networks

Isoform Function Prediction Based on Heterogeneous Graph Att...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Kuo Guo Yifan Li Hao Chen Hong-Bin Shen Yang Yang Department of Computer Science and Engineering Shanghai Jiao Tong University and Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai China Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University and Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai China Computational Biology Department School of Computer Science Carnegie Mellon University Pittsburgh PA USA

Isoforms refer to different mRNA molecules transcribed from the same gene, which can be translated into proteins with varying structures and functions. Predicting the functions of isoforms is an essential topic in bioinformatics as it can provide valuable insights into the intricate mechanisms of gene regulation and biological processes. Conventionally, gene function labels are standardized in Gene Ontology (GO) terms. However, traditional methods for predicting isoform function are largely limited by the absence of isoform-specific labels, sparse annotations, and the vast number of GO terms. To address these issues, we propose HANIso, a deep learning-based method for isoform function prediction. HANIso leverages a pretrained protein language model to extract features from protein sequences. It also integrates heterogeneous information, such as isoform sequence features, GO annotations, and isoform interaction data, using a Heterogeneous Graph Attention Network (HAN). This allows the model to learn the importance of different sources of information and their semantic relationships through the attention mechanism. Our method can predict function labels at both the gene level and isoform level. We conduct experiments on two species datasets, and the results demonstrate that our method outperforms existing methods on both AUROC and AUPRC. HANIso has the potential to overcome the limitations of traditional methods and provide a more accurate and comprehensive understanding of isoform function.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Online Discriminative Semantic-Preserving Hashing for Large-Scale Cross-Modal Retrieval 18th

Online Discriminative Semantic-Preserving Hashing for Large-...

引用

18th Pacific Rim International Conference on Artificial Intelligence, PRICAI 2021

作者： Yi, Jinhan He, Yi Liu, Xin Department of Computer Science and Technology Huaqiao University Xiamen361021 China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Nanjing University of Science and Technology Nanjing210094 China Xiamen Key Laboratory of Computer Vision and Pattern Recognition Xiamen China Fujian Key Laboratory of Big Data Intelligence and Security Xiamen China

ISBN: (纸本)9783030891879

Cross-modal hashing has drawn increasing attentions for efficient retrieval across different modalities, and existing methods primarily learn the hash functions in a batch based mode, i.e., offline methods. Nevertheless, the multimedia data often comes in a streaming fashion, which makes the batch based learning methods uncompetitive for large-scale streaming data due to the large memory consumption and calculation. To address this problem, we present an Online Discriminative Semantic-Preserving Hashing (ODSPH) method for large-scale cross-modal retrieval, featuring on fast training speed, low memory consumption and high retrieval accuracy. Within the proposed ODSPH framework, we utilize the newly coming data points to learn the hash codes and update hash functions in a stream manner. When new data comes, the corresponding hash codes are obtained by regressing the class label of the training examples. For hash function, we update it with the accumulated information from each round. Besides, we design a novel momentum updating method to adaptively update the hash function and reduce quantization loss, which can produce discriminative hash codes for high retrieval precision. Extensive experiments on three benchmark datasets show that the proposed ODSPH method improves the retrieval performance over the state-of-the-arts. © 2021, Springer Nature Switzerland AG.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Provable Discriminative Hyperspherical Embedding for Out-of-Distribution Detection 39

Provable Discriminative Hyperspherical Embedding for Out-of-...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Zou, Zhipeng Wan, Sheng Li, Guangyu Han, Bo Liu, Tongliang Zhao, Lin 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 Hong Kong Baptist University China Sydney AI Centre The University of Sydney Sydney Australia Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University China

ISBN: (纸本)157735897X

Out-of-distribution (OOD) detection aims to identify the test examples that do not belong to the distribution of training data. The distance-based methods, which identify OOD examples based on their distances from the centroids of in-distribution (ID) examples, have demonstrated promising OOD detection performance. However, the objectives utilized in prior approaches are typically designed for classification and thus might not yield sufficient discriminative power to distinguish between ID and OOD examples. Therefore, this paper proposes a prototype-based contrastive learning framework for OOD detection, which is termed provable Discriminative Hyperspherical Embedding (DHE). The proposed framework provides a theoretical analysis of inter-class dispersion, which is proved to be fundamental in reducing the false positive rate (FPR) on OOD examples. Based on this, we devise an angular spread loss to achieve the maximal dispersion of the prototypes of different classes prior to training. Subsequently, a prototype-enhanced contrastive loss is introduced to align embeddings of ID examples closely with their corresponding prototypes. In our proposed DHE, the maximal prototype dispersion is theoretically proved, thereby avoiding the pitfalls of local optima commonly encountered by most existing methods. Experimental results demonstrate the effectiveness of our proposed DHE, which showcases a remarkable reduction in FPR95 (i.e., 5.37% on CIFAR-100) and more than doubling the computational efficiency when compared with the state-of-the-art methods. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Contrastive Learning

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：