检索结果-内蒙古大学图书馆

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhu, Yongxin Li, Bocheng Zhang, Hang Li, Xin Xu, Linli Bing, Lidong School of Data Science University of Science and Technology of China China School of Computer Science and Technology University of Science and Technology of China China State Key Laboratory of Cognitive Intelligence China Zhejiang University China

Latent-based image generative models, such as Latent Diffusion Models (LDMs) and Mask Image Models (MIMs), have achieved notable success in image generation tasks. These models typically leverage reconstructive autoencoders like VQGAN or VAE to encode pixels into a more compact latent space and learn the data distribution in the latent space instead of directly from pixels. However, this practice raises a pertinent question: Is it truly the optimal choice? In response, we begin with an intriguing observation: despite sharing the same latent space, autoregressive models significantly lag behind LDMs and MIMs in image generation. This finding contrasts sharply with the field of NLP, where the autoregressive model GPT has established a commanding presence. To address this discrepancy, we introduce a unified perspective on the relationship between latent space and generative models, emphasizing the stability of latent space in image generative modeling. Furthermore, we propose a simple but effective discrete image tokenizer to stabilize the latent space for image generative modeling by applying K-Means on the latent features of self-supervised learning models. Experimental results show that image autoregressive modeling with our tokenizer (DiGIT) benefits both image understanding and image generation with the next token prediction principle, which is inherently straightforward for GPT models but challenging for other generative models. Remarkably, for the first time, a GPT-style autoregressive model for images outperforms LDMs, which also exhibits substantial improvement akin to GPT when scaling up model size. Our findings underscore the potential of an optimized latent space and the integration of discrete tokenization in advancing the capabilities of image generative models. The code is available at https://***/DAMO-NLP-SG/DiGIT. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Mapping the Scientific Landscape of Integrated Scheduling Towards Industry 5.0 with A Bibliometric Analysis 6

Mapping the Scientific Landscape of Integrated Scheduling To...

引用

6th International Conference on Industry 4.0 and Smart Manufacturing, ISM 2024

作者： Putri Ardyanti, Anak Agung Ayu Xie, Zhiqiang Karnyoto, Andrea Stevens Pardamean, Bens Estiyanti, Ni Made School of Computer Science and Technology Harbin University of Science and Technology Harbin150000 China Faculty of Information Technology and Design Primakara University Bali 80226 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program Master of Computer Science Bina Nusantara University Jakarta 11480 Indonesia School of Management and Business University of Debrecen Debrecen Hungary

The industry is rapidly transitioning from the 4.0 era to the 5.0 era, prompting renewed interest among scholars in scheduling problems. They allow operations to process and assemble various components simultaneously. This bibliometric analysis provides a comprehensive understanding of the diverse research perspectives on the potential challenges of integrated scheduling within the framework. We identified 357 Scopus articles and analysed co-occurrence keyword (CNK) by VOSViewer tools and Global Citation Score (GCS) by Scopus tools. The research data obtained from 1993 to 2023 mainly in the subject areas of Engineering, computer science, Maths, and others. CNK identified six cluster keywords, and nine most cited articles based on normalised GCS. These results contribute to our understanding of the development of scheduling algorithms, shifting from an emphasis on manufacturing transfers to contemporary demand-centric research directions. Furthermore, this research provides a valuable perspective for policymakers, plant personnel, and manufacturing managers to make informed decisions. © 2025 The Authors. Published by Elsevier B.V.

关键词： Job shop scheduling

来源：评论

学校读者我要写书评

暂无评论

Combining DistilBERT and LSTM to Enhanced Detection of Depression in Social Media Text 9

Combining DistilBERT and LSTM to Enhanced Detection of Depre...

引用

9th International Conference on Information Management and Technology, ICIMTech 2024

作者： Kurniadi, Felix Indra Satyaning Pradnya Paramita, Ni Luh Putu Angela Sihotang, Erna Fransisca Anggreainy, Maria Susan Zhan, Rongfang School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia School of Computer Science Bina Nusantara University Statistics Department Jakarta11480 Indonesia BINUS Graduate Program - Doctor of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia School of Computer Science University of North Texas Department of Rehabilitation and Health Service United States

ISBN: (纸本)9798350390025

Depression affects a significant number of people, and it has a significant impact not only on their lives but also on society as a whole. In light of this, we require more advanced methods that are capable of locating individuals who are depressed in a timely and accurate manner. This research investigates the effectiveness of Long Short-Term Mem- ory (LSTM) networks used in conjunction with DistilBERT, a small transformer-based model, in identifying indications of melancholy in posts made on social networking platforms. The text was cleaned up using a dataset that was obtained from Reddit, and then we used the mixed DistilBERT+LSTM model to evaluate how well it performed in comparison to the DistilBERT model by itself. The findings that we obtained indicate that both approaches produce results that are comparable but not flawless, with only slight variations in terms of accuracy, precision, recall, and F1 scores. The fact that this is the case demonstrates that there are issues with optimizing the model and displaying the data. © 2024 IEEE.

关键词： Social networking (online)

来源：评论

学校读者我要写书评

暂无评论

NEURON-ENHANCED AUTOENCODER MATRIX COMPLETION AND COLLABORATIVE FILTERING: THEORY AND PRACTICE 12

NEURON-ENHANCED AUTOENCODER MATRIX COMPLETION AND COLLABORAT...

引用

12th International Conference on Learning Representations, ICLR 2024

作者： Fan, Jicong Chen, Rui Zhang, Zhao Ding, Chris H.Q. School of Data Science The Chinese University of Hong Kong Shenzhen China Shenzhen Research Institute of Big Data Shenzhen China School of Computer Science & Information Engineering Hefei University of Technology China

Neural networks have shown promising performance in collaborative filtering and matrix completion but the theoretical analysis is limited and there is still room for improvement in terms of the accuracy of recovering missing values. This paper presents a neuron-enhanced autoencoder matrix completion (AEMC-NE) method and applies it to collaborative filtering. Our AEMC-NE adds an element-wise autoencoder to each output of the main autoencoder to enhance the reconstruction capability. Thus it can adaptively learn an activation function for the output layer to approximate possibly complicated response functions in real data. We provide theoretical analysis for AEMC-NE as well as AEMC to investigate the generalization ability of autoencoder and deep learning in matrix completion, considering both missing completely at random and missing not at random. We show that the element-wise neural network has the potential to reduce the generalization error bound, the data sparsity can be useful, and the prediction performance is closely related to the difference between the numbers of variables and samples. The numerical results on synthetic data and five benchmark datasets demonstrated the effectiveness of AEMC-NE in comparison to many baselines. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

FusionRender: Harnessing WebGPU's Power for Enhanced Graphics Performance on Web Browsers 24

FusionRender: Harnessing WebGPU's Power for Enhanced Graphic...

引用

33rd ACM Web Conference, WWW 2024

作者： Bi, Weichen Ma, Yun Han, Yudong Chen, Yifan Tian, Deyu Du, Jiaqi Institute for Artificial Intelligence Peking University Beijing China College of Computer and Data Science Fuzhou University Fuzhou China School of Computer Science Peking University Beijing China School of Electronics Engineering and Computer Science Peking University Beijing China

ISBN: (纸本)9798400701719

Graphics rendering on web browsers serves as the foundation for numerous web applications. Compared with the widely employed WebGL, the next-generation web graphics API, WebGPU, demonstrates an enhanced capacity to adapt to modern GPU features, boasting more significant potential. However, our experiment shows that the performance of current graphics rendering frameworks based on WebGPU lags behind those built on WebGL. Such discrepancy primarily arises from an incomplete alignment with WebGPU's distinctive features. The individual rendering of each graphic leads to redundant communication between the CPU and GPU. To enhance the graphics performance on the web, we introduce the FusionRender to harness the power of WebGPU. To mitigate redundant communication, FusionRender assigns a unique signature to each object and employs these signatures for grouping, enabling the consolidation of graphics rendering whenever possible. In simulated experiments involving the rendering of multiple objects, FusionRender improves the rendering performance by 29.3%-122.1% compared with the existing optimal baseline. In real cases with more complex features, performance improvement ranges from 9.4% to 39.7%. Additionally, FusionRender exhibits robust performance enhancement across various devices and browsers. © 2024 ACM.

关键词： Web browsers

来源：评论

学校读者我要写书评

暂无评论

Indonesian Agricultural-crops Classification Using Transfer Learning Model 8

Indonesian Agricultural-crops Classification Using Transfer ...

引用

8th International Conference on computer science and Computational Intelligence, ICCSCI 2023

作者： Isnan, Mahmud Hidayat, Alam Ahmad Pardamean, Bens Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia Department of Mathematics School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program - Master of Computer Science Program Bina Nusantara University Jakarta11480 Indonesia

Classification of Indonesian crops is a critical task in developing farming and getting more understanding of agriculture. However, there is no clear task in classifying types of crops in Indonesia. Transfer learning has been used successfully in a variety of image classification applications. Thus, in this paper, we collected images of Indonesian crops from the internet randomly and proposed a classification by using transfer learning of deep learning with four pre-trained models: EffficientNetB0, ResNet18, VGG19, and AlexNet. In the experiment, augmentation techniques such as random horizontal flip, random vertical flip, and random affine were utilized to prevent the network from overfitting. The result found that EfficientNet-B0 outperformed other models with an accuracy of 82.55. Then, the model struggled to distinguish between crops in the same family. According to the results, although transfer learning can work well to classify images of Indonesian agricultural crops, some improvements are still required to address existing issues. © 2023 The Authors. Published by Elsevier B.V.

关键词： Agriculture Classification Crop Deep Learning Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Security Performance Analysis for MIMO Systems: From Artificial Noise's Power and Dimension Perspective

引用

IEEE Transactions on Vehicular Technology 2025年第6期74卷 9985-9989页

作者： Xu, Panpan Li, Zhenling Li, Chunguo Tan, Weijie Guizhou University School of Mathematics and Statistics Guiyang550025 China Southeast University School of Information Science and Engineering Nanjing210096 China Guizhou University State Key Laboratory of Public Big Data College of Computer Science and Technology Guiyang550025 China

This paper analyzes the influence of power and dimension of artificial noise (AN) on security performance of multiple-input multiple-output (MIMO) system with multiple randomly located eavesdroppers. We derive the closed-form expressions of the secrecy outage probability (SOP) and the optimal secrecy capacity by utilizing the irrelevance among transmission channels, and the orthogonality between AN and the basis vectors of the legitimate channel matrix's null space. We use the Poisson point process to determine the number and location of eavesdroppers, then analyze the security performance of the communication system under the 3D channel condition of probability-based line-of-sight (LoS) or non-line-of-sight (NLoS) channel. The simulation results show the following conclusions. When the AN dimension reaches 1/4 highest dimension (i.e., one quarter of the number of transmitting antennae less than one) and the power splitter factor (PSF) falls in the open interval (0.6, 0.9), we can achieve a trade-off between secrecy capacity and design complexity. © 1967-2012 IEEE.

关键词： Vector spaces

来源：评论

学校读者我要写书评

暂无评论

Expressive Voice Conversion Using Perturbation and Autoencoder Loss 12

Expressive Voice Conversion Using Perturbation and Autoencod...

引用

12th IEEE International Conference on Information and Communication Networks, ICICN 2024

作者： Chen, Zhe Liu, Yining Liu, Gao School of Computer Science and Information Security Guilin University of Electronic Technology Guilin China School of Data Science and Artificial Intelligence Wenzhou University of Technology Wenzhou China College of Computer Science Chongqing University Chongqing China

ISBN: (纸本)9798350355802

Expressive voice conversion is to transform voice speech with emotional expression. Modeling emotional styles in expressive voice conversion is challenging. It is essential to eliminate the interference of factors such as the identity in emotional style modeling. This paper addresses this issue using an information perturbation method. During the feature extraction phase, perturbation functions are applied to factors such as formants and pitch in the speech signal to eliminate the interference of speaker information. This allows for separate modeling and extraction of content information and emotional information. We propose a general framework for expressive voice conversion, which aims to disentangle content information, speaker information, and emotional style information from the voice signal. Experimental results demonstrate the effectiveness of our proposed framework in both subjective and objective evaluations. It can extract highly expressive and understandable features from the source speech while maintaining similarity to the target speaker. © 2024 IEEE.

关键词： Signal modulation

来源：评论

学校读者我要写书评

暂无评论

Count-data Mixed Models of Topical Tweets: A Case of Indonesia Flood Events 8

Count-Data Mixed Models of Topical Tweets: A Case of Indones...

引用

8th International Conference on computer science and Computational Intelligence, ICCSCI 2023

作者： Hidayat, Alam Ahmad Pardamean, Bens Mathematics Department School of Computer Science Bina Nusantara University Jakarta11480 Indonesia Bioinformatics and Data Science Research Center Bina Nusantara University Jakarta11480 Indonesia Computer Science Department BINUS Graduate Program - Master of Computer Science Program Bina Nusantara University Jakarta11480 Indonesia

Examining topic-level variability in modeling Twitter data can potentially yield more comprehensive insights into public perception during critical periods, thereby enhancing natural disaster mitigation and surveillance efforts. In this study, we utilized generalized linear mixed models (GLMMs) to illustrate the variability in tweet counts related to specific topics in Indonesia during the flood events that occurred in February 2021. The glmmTMB library in R was employed for this purpose. The data were assumed to follow two distinct exponential distributions: Poisson and Negative Binomial. To incorporate random effects, random intercepts and random slopes were introduced, allowing them to vary randomly across topics in the initial two models. Additionally, the final model addressed issues related to dispersion and zero-inflation. By evaluating the Akaike Information Criteria scores, we determined that a model based on the Negative Binomial distribution with random zero-inflation intercepts best fit the data. The chosen model formulation and the estimated parameters have the potential to forecast topic-specific trends in Indonesian flood-related Twitter data. © 2023 The Authors. Published by Elsevier B.V.

关键词： count data disaster surveillance floods mixed models Twitter

来源：评论

学校读者我要写书评

暂无评论

Application of improved virtual sample and sparse representation in face recognition

引用

CAAI Transactions on Intelligence Technology 2023年第4期8卷 1391-1402页

作者： Yongjun Zhang Zewei Wang Xuexue Zhang Zhongwei Cui Bob Zhang Jinrong Cui Lamin LJanneh College of Computer Science and Technology Guizhou UniversityGuiyangChina Big Data Science and Intelligent Engineering Research Institute Guizhou Education UniversityGuiyangChina School of Mathematics and Big Data Guizhou Education UniversityGuiyangChina Department of Computer and Information Science University of MacaoAvenida da UniversidadeTaipaMacaoChina School of Mathematics and Information South China Agricultural UniversityGuangzhouChina

Sparse representation plays an important role in the research of face *** a deformable sample classification task,face recognition is often used to test the performance of classification *** face recognition,differences in expression,angle,posture,and lighting conditions have become key factors that affect recognition ***,there may be significant differences between different image samples of the same face,which makes image classification very ***,how to build a robust virtual image representation becomes a vital *** solve the above problems,this paper proposes a novel image classification ***,to better retain the global features and contour information of the original sample,the algorithm uses an improved non‐linear image representation method to highlight the low‐intensity and high‐intensity pixels of the original training sample,thus generating a virtual ***,by the principle of sparse representation,the linear expression coefficients of the original sample and the virtual sample can be calculated,*** obtaining these two types of coefficients,calculate the distances between the original sample and the test sample and the distance between the virtual sample and the test *** two distances are converted into distance ***,a simple and effective weight fusion scheme is adopted to fuse the classification scores of the original image and the virtual *** fused score will determine the final classification *** experimental results show that the proposed method outperforms other typical sparse representation classification methods.

关键词： representation sample image

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：