检索结果-内蒙古大学图书馆

2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022

作者： Wang, Ya Sun, Xingwu Lian, Fengzong Kang, Zhanhui Xu, Chengzhong Machine Learning Platform Department Tencent State Key Lab of IOTSC Department of Computer Science University of Macau China

Position Embedding (PE) is essential for transformer to capture the sequence ordering of input tokens. Despite its general effectiveness verified in Natural Language Processing (NLP) and computer Vision (CV), its application in cross-modal tasks remains unexplored and suffers from two challenges: 1) the input text tokens and image patches are not aligned;2) the encoding space of each modality is different, making it unavailable for feature comparison. In this paper, we propose a unified position embedding method for these problems, called AnChor-basEd Relative Position Embedding (ACE-RPE), in which we first introduce an anchor locating mechanism to bridge the semantic gap and locate anchors from different modalities. Then we conduct the distance calculation of each text token and image patch by computing their shortest paths from the located anchors. Last, we embed the anchor-based distance to guide the computation of cross-attention. In this way, it calculates cross-modal relative position embeddings for cross-modal transformer. Benefiting from ACE-RPE, our method obtains new SOTA results on a wide range of benchmarks, such as Image-Text Retrieval on MS-COCO and Flickr30K, Visual Entailment on SNLI-VE, Visual Reasoning on NLVR2 and Weakly-supervised Visual Grounding on RefCOCO+. © 2022 Association for Computational Linguistics.

关键词： Anchors

来源：评论

学校读者我要写书评

暂无评论

Deep learning-Based Age and Gender Prediction through Feature Extraction from Facial Images using Convolutional Neural Networks 5

Deep Learning-Based Age and Gender Prediction through Featur...

引用

5th IEEE International Conference on Communication, Computing and Industry 6.0, C2I6 2024

作者： Manasa, S.M. Ramamoorthy, Raghu Velu, Anitha Angel Donny, F. Chaitrashree, S. Ramaiah Institute of Technology Department of Artificial Intelligence and Machine Learning Bengaluru India The Oxford College of Engineering Department of Computer Science and Engineering Bengaluru India Sri Sairam College of Engineering Department of Electronics and Communication Engineering Bengaluru India

ISBN: (纸本)9798350379884

In recent years, age estimation and gender classification have been focal points in the field of pattern recognition and computer vision. This paper introduces an automated approach for predicting age and gender by extracting features from human facial images. In contrast to conventional methods that operate on unfiltered facial images, our study demonstrates significant improvements in these tasks through the use of deep convolutional neural networks (CNNs). The employed feedforward neural network method enhances robustness for highly variable unconstrained recognition tasks, specifically targeting gender identification and age group estimation. The research methodology involves the analysis and validation of gender prediction and age estimation on two benchmark datasets: the Essex face dataset and the Adience benchmark dataset. Results indicate a substantial performance enhancement, showcasing the proposed approach's efficiency and state-of-the-art performance in both age and gender detection. The model achieves remarkable accuracy, positioning it as a notable advancement in the realm of age and gender prediction using deep learning techniques. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Scheduling the Tasks in Cloud Computing using Adaptive Adjustment in Whale Optimization Algorithm 2

Scheduling the Tasks in Cloud Computing using Adaptive Adjus...

引用

2nd International Conference on Data science and Information System, ICDSIS 2024

作者： Pareek, Piyush Kumar Nithyashree, G.D. Rajashekhar, S.A. Zanke, Pankaj Navyashree, K.S. Nitte Meenakshi Institute Of Technology Artificial Intelligence And Machine Learning Bengaluru India East West Institute Of Technology Department Of Computer Science And Engineering Bengaluru India Sapiens International Corporation United States

ISBN: (纸本)9798350381665

One of the most popular infrastructures for employing virtual machines, also known as processing units, is cloud computing. Task scheduling is fundamental problems of cloud computing. A number of methods to handle the nondeterministic polynomial-time (NP)-complete optimisation issue of choosing the best scheduling criteria in cloud computing have been proposed by academics in the past. An effective optimisation approach is suggested in this research, based on the whale optimisation algorithm (WOA), to address the common problems of solution algorithms, which include poor global optimisation ability, ease of sliding into local optimisation, and sluggish convergence speed. We provide an adaptive adjustment method that may dynamically alter search behaviour during the whale optimisation algorithm's iterative operation. The technique described to further update the position of search individuals also uses the evolutionary approach of differential variation. The algorithm's exploration and development capacities are then balanced, and a restart approach is included to keep the system from entering local optimisation. The approach simultaneously resolves the work planning issue in the situation of rain cloud processing. When compared to other algorithms, this technique uses more resources and converges more quickly. Additionally, the evaluation function seeks for the optimal answers by taking the makespan, resource cost, and load balancing level into account. © 2024 IEEE.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

UNDERSTANDING THE GENERALIZATION OF ADAM IN learning NEURAL NETWORKS WITH PROPER REGULARIZATION 11

UNDERSTANDING THE GENERALIZATION OF ADAM IN LEARNING NEURAL ...

引用

11th International Conference on learning Representations, ICLR 2023

作者： Zou, Difan Cao, Yuan Li, Yuanzhi Gu, Quanquan Department of Computer Science Institute of Data Science The University of Hong Kong Hong Kong Department of Statistics & Actuarial Science The University of Hong Kong Hong Kong Machine Learning Department Carnegie Mellon University United States Department of Computer Science University of California Los Angeles United States

Adaptive gradient methods such as Adam have gained increasing popularity in deep learning optimization. However, it has been observed in many deep learning applications such as image classification, Adam can converge to a different solution with a worse test error compared to (stochastic) gradient descent, even with a fine-tuned regularization. In this paper, we provide a theoretical explanation for this phenomenon: we show that in the nonconvex setting of learning over-parameterized two-layer convolutional neural networks starting from the same random initialization, for a class of data distributions (inspired from image data), Adam and gradient descent (GD) can converge to different global solutions of the training objective with provably different generalization errors, even with weight decay regularization. In contrast, we show that if the training objective is convex, and the weight decay regularization is employed, any optimization algorithms including Adam and GD will converge to the same solution if the training is successful. This suggests that the generalization gap between Adam and SGD in the presence of weight decay regularization is closely tied to the nonconvex landscape of deep learning optimization, which cannot be covered by the recent neural tangent kernel (NTK) based analysis. © 2023 11th International Conference on learning Representations, ICLR 2023. All rights reserved.

关键词： Gradient methods

来源：评论

学校读者我要写书评

暂无评论

From 5G to 6G: The Role of AI, machine learning, and Deep learning in Wireless Systems 4

From 5G to 6G: The Role of AI, Machine Learning, and Deep Le...

引用

4th International Conference on Sentiment Analysis and Deep learning, ICSADL 2025

作者： Gote, Pradnyawant M. Kumar, Praveen Verma, Prateek Yesankar, Prajyot Pawar, Adesh Saratkar, Saniya Faculty of Engineering and Technology Department of Computer Science & Design Maharashtra Wardha442001 India Faculty of Engineering and Technology Department of Computer Science & Medical Engineering Maharashtra Wardha442001 India Faculty of Engineering and Technology Department of Artificial Intelligence & Machine Learning Maharashtra Wardha442001 India Faculty of Engineering and Technology Department of Artificial Intelligence & Data Science Maharashtra Wardha442001 India

ISBN: (纸本)9798331523923

The swift progression of wireless communication technologies-specifically from 5G to 6G is an approach that could be the most significant revolutionary leap towards changing connectivity and data transmission forever. Additionally, algorithms of machine learning (ML) and deep learning (DL) in 5G and 6G can offer promising advancements toward predictive maintenance, dynamic spectrum management, and enhanced security. The review of the potentially transformative nature of these technologies underlines the need for future directions of research efforts called for in circumventing challenges towards making full capability exploitation of 6G possible. © 2025 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

EPR-Net: constructing a non-equilibrium potential landscape via a variational force projection formulation

引用

National science Review 2024年第7期11卷 135-147页

作者： Yue Zhao Wei Zhang Tiejun Li Center for Data Science Peking University Zuse Institute Berlin Department of Mathematics and Computer Science Freie Universit?t Berlin Laboratory of Mathematics and Applied Mathematics(LMAM)and School of Mathematical Sciences Peking University Center for Machine Learning Research Peking University

We present EPR-Net, a novel and effective deep learning approach that tackles a crucial challenge in biophysics: constructing potential landscapes for high-dimensional non-equilibrium steady-state ***-Net leverages a nice mathematical fact that the desired negative potential gradient is simply the orthogonal projection of the driving force of the underlying dynamics in a weighted inner-product ***, our loss function has an intimate connection with the steady entropy production rate(EPR),enabling simultaneous landscape construction and EPR estimation. We introduce an enhanced learning strategy for systems with small noise, and extend our framework to include dimensionality reduction and the state-dependent diffusion coefficient case in a unified fashion. Comparative evaluations on benchmark problems demonstrate the superior accuracy, effectiveness and robustness of EPR-Net compared to existing methods. We apply our approach to challenging biophysical problems, such as an eight-dimensional(8D)limit cycle and a 52D multi-stability problem, which provide accurate solutions and interesting insights on constructed landscapes. With its versatility and power, EPR-Net offers a promising solution for diverse landscape construction problems in biophysics.

关键词： high-dimensional potential landscape non-equilibrium system entropy production rate dimensionality reduction deep learning

来源：评论

学校读者我要写书评

暂无评论

Home Automation Using Wi-Fi: ESP32-Based System for Remote Control and Environmental Monitoring 3

Home Automation Using Wi-Fi: ESP32-Based System for Remote C...

引用

3rd IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics, ICDCECE 2024

作者： Abdul Lateef Haroon, P.S. Shafiulla, Mohammed Naveed, Shaik Mohammed Ahmed, Suhail Nawaz, Shaik Mohammed Kumar, Utkarsh Ballari Institute Of Technology & Management Department Of Electronics And Communication Engineering Ballari India Ballari Institute Of Technology And Management Department Of Computer Science And Engineering Ballari India Ballari Institute Of Technology & Management Department Of Computer Science And Engineering Ballari India Ballari Institute Of Technology & Management Department Of Artificial Intelligence And Machine Learning Ballari India

ISBN: (纸本)9798350318609

This project introduces a comprehensive home automation system leveraging Wi-Fi connectivity and ESP32 micro controllers, aimed at enhancing modern living through seamless automation and user-centric design. The background emphasizes the fusion of IoT technology with home automation, highlighting Wi-Fi's significance in enabling remote monitoring and control of household devices. Methodologically, the project integrates relay modules and DHT11 sensors, orchestrated through a user-friendly smartphone application. Results demonstrate successful implementation and testing, showcasing functionality, reliability, and user-friendliness. Real-time environmental monitoring and remote appliance control fulfill the objectives of enhancing home comfort, convenience, and energy efficiency. The conclusion underscores the project's contribution to advancing smart home technologies, offering users a scalable and customizable solution. Future research opportunities include enhancing system capabilities and addressing security and privacy concerns. © 2024 IEEE.

关键词： Microcontrollers

来源：评论

学校读者我要写书评

暂无评论

Digital Evidence Management: Modern Techniques in Digital Forensics 3

Digital Evidence Management: Modern Techniques in Digital Fo...

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Venkataramanan, N. Jayalakshmi, V. Anitha, S. Ganesan, R. Jenifer, V. Drbccc Hindu College Department of Computer Applications Pattabiram Chennai600072 India Department of Computer Science and Applications Ramapuram600089 India Anna University St. Joseph's College of Engineering Department of Artificial Intelligence and Machine Learning Chennai600119 India Vel Tech Rangarajan Dr.Sagunthala R&d Institute of Science and Technology Department of Computer Science and Engineering Chennai600062 India Department of Computer Science and Applications Ramapuram Campus 600089 India

ISBN: (纸本)9798331509675

This research study reviews the significance of digital forensic analysis in modern criminal investigations, highlighting the way it complements conventional methods of inquiry and the difficulties it encounters in the information era. To solve crimes effectively, investigators must combine digital knowledge with traditional forensic techniques. This will enable them to use evidence from many different types of digital and physical sources. The specific intricacies of digital crime scenes need specialized methods for the proper storage and analysis of digital evidence to identify criminal intent and patterns of conduct. To propose Modern Techniques for Digital Evidence Management [MT-DEM] based Digital Forensics Service. Multidisciplinary teams and cutting-edge technology are becoming necessities in digital forensic investigation as the field adjusts to new threats like data encryption and massive amounts of data. The admissibility of evidence in the legal world depends on digital forensic procedures being transparent and adhering to set criteria. Digital forensic analysis has become an indispensable tool for contemporary investigators with the rise of digital evidence. © 2025 IEEE.

关键词： Forensic science

来源：评论

学校读者我要写书评

暂无评论

Low-Rank Optimal Transport through Factor Relaxation with Latent Coupling

arXiv

引用

arXiv 2024年

作者： Halmos, Peter Liu, Xinhao Gold, Julian Raphael, Benjamin J. Department of Computer Science Princeton University United States Center for Statistics and Machine Learning Princeton University United States

Optimal transport (OT) is a general framework for finding a minimum-cost transport plan, or coupling, between probability distributions, and has many applications in machine learning. A key challenge in applying OT to massive datasets is the quadratic scaling of the coupling matrix with the size of the dataset. Forrow et al. (2019) introduced a factored coupling for the k-Wasserstein barycenter problem, which Scetbon et al. (2021) adapted to solve the primal low-rank OT problem. We derive an alternative parameterization of the low-rank problem based on the latent coupling (LC) factorization previously introduced by Lin et al. (2021) generalizing Forrow et al. (2019). The LC factorization has multiple advantages for low-rank OT including decoupling the problem into three OT problems and greater flexibility and interpretability. We leverage these advantages to derive a new algorithm Factor Relaxation with Latent Coupling (FRLC), which uses coordinate mirror descent to compute the LC factorization. FRLC handles multiple OT objectives (Wasserstein, Gromov-Wasserstein, Fused Gromov-Wasserstein), and marginal constraints (balanced, unbalanced, and semi-relaxed) with linear space complexity. We provide theoretical results on FRLC, and demonstrate superior performance on diverse applications – including graph clustering and spatial transcriptomics – while demonstrating its interpretability. Copyright © 2024, The Authors. All rights reserved.

关键词： Matrix factorization

来源：评论

学校读者我要写书评

暂无评论

DBROA based Convolutional Auto Encoder for European Credit Card Fraud Detection 2

DBROA based Convolutional Auto Encoder for European Credit C...

引用

2nd International Conference on Data science and Information System, ICDSIS 2024

作者： Pareek, Piyush Kumar Mohan, B.R. Vidya, E.V. Zanke, Pankaj Navyashree, K.S. Nitte Meenakshi Institute Of Technology Artificial Intelligence And Machine Learning Bengaluru India East West Institute Of Technology Department Of Computer Science And Engineering Bengaluru India Sapiens International Corporation United States

ISBN: (纸本)9798350381665

Numerous studies using bad neuronic nets (DNNs) for the objective of detecting recognition card fraud have concentrated on enhancing point prediction accuracy and limiting unwelcome biases by developing various network designs or learning models. It is crucial to reckon uncertainty together with point estimate of the training data has a significant impact on how well machine learning algorithms function, therefore data imbalance is not a minor problem. In general, the data only shows a tiny portion of false communications. The effectiveness of simple machine acquiring classifiers is significantly impacted by this. This research studies several data augmentation strategies to handle the unbalanced data problem for credit card fraud detection in order to deal with the shortage of false events. The balanced data is then sent to a Convolutional Auto Encoder (CAE) for classification, and a Developed Battle Royale optimisation algorithm (DBROA) is used to determine the model's weight in the best possible way. For the purpose of recognising credit card fraud involving European cardholders, each model includes stratified K-fold cross-validation and a real-world dataset. The results show that the suggested model is more accurate than current models. © 2024 IEEE.

关键词： Convolution

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：