检索结果-内蒙古大学图书馆

The Distributed Intelligent Collaboration to UAV-Assisted VEC: Joint Position Optimization and Task Scheduling

IEEE Internet of Things Journal 2025年第12期12卷 21473-21487页

作者： Yi, Meng Lee, Vincent C.S. Yang, Peng Li, Peisong Zhang, Yifan Wei, Wei Gao, Honghao Southeast University School of Computer Science and Engineering Nanjing211189 China Monash University Faculty of IT Department of DS and AI ClaytonVIC3800 Australia Southeast University School of Computer Science and Engineering the Key Laboratory of Computer Network and Information Integration Ministry of Education Nanjing211189 China Xidian University Hangzhou Institute of Technology Hangzhou311200 China Southeast University-Monash University Joint Graduate School Suzhou215123 China Xi'an University of Technology School of Computer Science and Engineering Xi'an710048 China Shanghai University School of Computer Engineering and Science Shanghai200444 China

Deploying Unmanned Aerial Vehicles (UAVs) as aerial base stations enhances the coverage and performance of communication networks in Vehicular Edge Computing (VEC) scenarios. However, due to the limited communication range and energy capacity of UAVs, they cannot continuously cover entire areas or sustain long flights. Therefore, achieving full communication coverage of a target area with a minimal number of UAVs and efficient task offloading remains a significant challenge. To address this problem, the UAV-assisted Two-stage Intelligent Collaboration (UTIC) method is proposed in this paper to tackle the joint position optimization and task scheduling issue. Firstly, a UAV-assisted Two-stage Task Scheduling (UTTS) system model is designed to optimize the allocation process. Secondly, an Enhanced Particle Swarm Optimization (E-PSO) algorithm is designed to determine the optimal positions of UAVs, ensuring complete coverage of all mobile vehicles (MVs) with the minimum number of UAVs. Thirdly, Deep Deterministic Policy Gradient (DDPG) method is employed to find the optimal scheduling decisions for MVs, considering energy consumption, delay, and task priorities. Simulation results demonstrate that the proposed UTIC method can achieve nearly 20% reduction in UAV deployment and outperform three other classical reinforcement learning (RL) algorithms in terms of reducing system cost. © 2025 IEEE.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

Detection of Unathorized Vechile in BRT Lane

Detection of Unathorized Vechile in BRT Lane

引用

IEEE Punecon

作者： Pallavi Nikumbh Pankaj More Harsh Vahal Pratham Mali Prathmesh Mohite Department of Computer Science and Engineering (AI&ML) Pimpri Chinchwad College of Engineering Pune India Department of Computer Science and Engineering Pimpri Chinchwad College of Engineering Pune India

ISBN: (数字)9798331527822

ISBN: (纸本)9798331527839

The increasing number of unauthorized vehicles in the Bus Rapid Transit (BRT) lanes creates a significant challenge to urban transportation. Due to this efficiency and quality of safety decreases. Our project aims to develop an automated system that detects unauthorized vehicles that come into BRT lanes using deep learning techniques like YOLOv8. The system employs combination video surveillance, image processing, and deep learning models to accurately identify and track vehicles that violate the BRT regulations. The proposed solution enhances traffic management and reduces the incidence of unauthorized lane usage by integrating the real- time data processing and alert mechanisms. The system employs a dataset from Roboflow, extended with a self-made dataset containing images of various vehicles in BRT lanes. We utilize YOLOv8 for object detection to accurately identify and classify vehicles. Also, Easy-OCR is used to recognize text from license plates for validation and verification against authorized vehicle list. Our preliminary results demonstrate that the system achieves 80% accuracy in detecting unauthorized vehicles. Our solution is urban traffic management by reducing illegal lane usage and ensuring easier public transport operations.

关键词： Deep learning Text recognition Urban areas Transportation Object detection Video surveillance Regulation Safety Object recognition License plate recognition

来源：评论

学校读者我要写书评

暂无评论

Noise Contrastive Alignment of Language Models with Explicit Rewards 38

Noise Contrastive Alignment of Language Models with Explicit...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Chen, Huayu He, Guande Yuan, Lifan Cui, Ganqu Su, Hang Zhu, Jun Department of Computer Science and Technology Tsinghua University China Institute for AI BNRist Center Tsinghua-Bosch Joint ML Center THBI Lab Tsinghua University China Zhongguancun Laboratory Beijing China

User intentions are typically formalized as evaluation rewards to be maximized when fine-tuning language models (LMs). Existing alignment methods, such as Direct Preference Optimization (DPO), are mainly tailored for pairwise preference data where rewards are implicitly defined rather than explicitly given. In this paper, we introduce a general framework for LM alignment, leveraging Noise Contrastive Estimation (NCE) to bridge the gap in handling reward datasets explicitly annotated with scalar evaluations. Our framework comprises two parallel algorithms, NCA and InfoNCA, both enabling the direct extraction of an LM policy from reward data as well as preference data. Notably, we show that the DPO loss is a special case of our proposed InfoNCA objective under pairwise preference settings, thereby integrating and extending current alignment theories. By comparing NCA and InfoNCA, we demonstrate that the well-observed decreasing-likelihood trend of DPO/InfoNCA is caused by their focus on adjusting relative likelihood across different responses. In contrast, NCA optimizes the absolute likelihood for each response, thereby effectively preventing the chosen likelihood from decreasing. We evaluate our methods in both reward and preference settings with Mistral-8×7B and 7B models. Experiments suggest that InfoNCA/NCA surpasses various preference baselines when reward datasets are available. We also find NCA significantly outperforms DPO in complex reasoning tasks like math and coding. Code: https://***/thu-ml/Noise-Contrastive-Alignment. © 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

HOW ROBUST IS UNSUPERVISED REPRESENTATION LEARNING TO DISTRIBUTION SHIFT? 11

HOW ROBUST IS UNSUPERVISED REPRESENTATION LEARNING TO DISTRI...

引用

11th International Conference on Learning Representations, ICLR 2023

作者： Shi, Yuge Daunhawer, Imant Vogt, Julia E. Torr, Philip H.S. Sanyal, Amartya Department of Engineering Science University of Oxford United Kingdom Department of Computer Science ETH Zurich Switzerland Department of Computer Science ETH AI Center ETH Zurich Switzerland

The robustness of machine learning algorithms to distributions shift is primarily discussed in the context of supervised learning (SL). As such, there is a lack of insight on the robustness of the representations learned from unsupervised methods, such as self-supervised learning (SSL) and auto-encoder based algorithms (AE), to distribution shift. We posit that the input-driven objectives of unsupervised algorithms lead to representations that are more robust to distribution shift than the target-driven objective of SL. We verify this by extensively evaluating the performance of SSL and AE on both synthetic and realistic distribution shift datasets. Following observations that the linear layer used for classification itself can be susceptible to spurious correlations, we evaluate the representations using a linear head trained on a small amount of out-of-distribution (OOD) data, to isolate the robustness of the learned representations from that of the linear head. We also develop "controllable" versions of existing realistic domain generalisation datasets with adjustable degrees of distribution shifts. This allows us to study the robustness of different learning algorithms under versatile yet realistic distribution shift conditions. Our experiments show that representations learned from unsupervised learning algorithms generalise better than SL under a wide variety of extreme as well as realistic distribution shifts. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Analysis of Word Sense Disambiguation in a Regional Language

A Comprehensive Analysis of Word Sense Disambiguation in a R...

引用

International Conference on Inventive Computation Technologies (ICICT)

作者： Aditi Barai Manami Das Priyanka Bhowmick Udita Dey Prianka Dey Sagarika Chowdhury Department of Computer Science and Business Systems Narula Institute of Technology Kolkata India Department of Computer Science and Engineering (AI ML) Narula Institute of Technology Kolkata India

ISBN: (数字)9798331512248

ISBN: (纸本)9798331512255

Word Sense Disambiguation(WSD) is a Natural Language Processing(NLP) technique that tries to disambiguate ambiguous words by finding the right sense of a word in a particular context. It is a classification task that assists in mapping a word to one or more potential senses based on its sentence. Understanding the intended sense in a particular context is crucial for proper communication and accurate language processing. But WSD is especially crucial in Bengali for a number of reasons: Similar to other Indic languages, Bengali possesses a rich and complex system of word formation. A root word can be modified according to aspects like tense, gender, number, and case. An attempt is made in this working paper where authors will be working on WSD but on the Bengali language, where words have different meanings. The Bengali language has plenty of words with multiple meanings. Even though WSD has been propagated the world over, the language Bengali has a lot of challenges since there are very few resources available and research due to its complex syllable structure.

关键词： Surveys Sensitivity Computational modeling Evaluation models Natural languages Computational efficiency Complexity theory Object recognition Optimization Overfitting

来源：评论

学校读者我要写书评

暂无评论

Weather Forecasting Using Advanced Artificial Intelligence Techniques Based on Swin Transformer Network 2

Weather Forecasting Using Advanced Artificial Intelligence T...

引用

2nd International Conference on Distributed Computing and Optimization Techniques, ICDCOT 2024

作者： Kavitha, S. Nagamani, H.S. Sumanth, S. Pareek, Piyush Kumar Mageswari, P. Uma Nitte Meenakshi Institute of Technology Department of Mca Bangalore India Govt. Smt Vhd Central Institute of Home Science Department of Computer Science Bengaluru India Government College for Women Department of Computer Science and Applictaions Kolar India Nitte Meenakshi Institute of Technology Department of AI/ML Bangalore India Presidency College Department of Computer Applications Bangalore India

ISBN: (纸本)9798350382952

It is still difficult for weather prediction (NWP) representations to exactly anticipate rainfall, especially the highest rainfall events. The severely intermittent character rainfall, together with subgrid-scale are critical to the multi-scale dynamics that produce rainfall, might be the reasons for this. Because of this inherent fallibility of weather prediction model projections, it is standard practice to fix these forecasts by post-processing the model's output. To improve weather prediction, the study suggests using a Spectral Swin Transformer Network (SSTN). An operating high-resolution model system, the PRUFS model, which runs for East China at 3 km grid intermissions, was used to test the model. The 311 conventional ground-based weather stations to collect data on 2 m temperature, 2 m relative speed. A year's worth of historical PRUFS perfect outputs and surface comments make up the training dataset for SSTN, while two real-time SSTN outputs form the basis of the act evaluation. © 2024 IEEE.

关键词： Weather forecasting

来源：评论

学校读者我要写书评

暂无评论

Virtual-event-based posterior sampling and inference for Neyman-Scott processes

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2024年第1期25卷 18784-18850页

作者： Chengkuan Hong Christian R. Shelton Jun Zhu Department of Computer Science and Technology BNRist Center Tsinghua Institute for AI Tsinghua-Bosch Joint Center for ML Tsinghua University Beijing China Department of Computer Science and Engineering University of California Riverside CA

Neyman-Scott processes (NSPs) are a class of Cox processes constructed by stacking layers of Poisson processes into a deep structure. While a lot of research has been conducted regarding the posterior sampling and inference for NSPs, most of the existing methods only work for shallow NSPs (i.e., NSPs with one layer of latent Poisson processes). In this paper, we present virtual-event-based posterior sampling and inference algorithms for NSPs. The algorithms work for both deep NSPs and shallow NSPs. Moreover, we show that deep NSPs can be viewed as branching processes or a limiting case of probabilistic graphical models. We conduct a theoretical analysis of the convergence of our algorithms and provide the condition for the convergence to hold. In doing so, we also prove the convergence of virtual-event-based sampling inference algorithms for other point process models with missing information (Markov jump processes, piecewise-constant intensity models, and Hawkes processes). Like NSPs, the latent variables of these models with missing information are also point processes. Our experimental results demonstrate that the prediction based on our sampling and inference algorithms for NSPs can achieve good prediction performance compared with state-of-the-art methods.

关键词： Markov chain Monte Carlo variational inference point processes hierarchical model Neyman-Scott processes

来源：评论

学校读者我要写书评

暂无评论

CM-Off-Meme: Code-Mixed Hindi-English Offensive Meme Detection with Multi-Task Learning by Leveraging Contextual Knowledge 30

CM-Off-Meme: Code-Mixed Hindi-English Offensive Meme Detecti...

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Kumari, Gitanjali Bandyopadhyay, Dibyanayan Ekbal, Asif Vinutha, B.N. Department of Computer Science and Engineering Indian Institute of Technology Patna Bihta India Wipro AI Labs Karnataka Bangalore India

ISBN: (纸本)9782493814104

Detecting offensive content in internet memes is challenging as it needs additional contextual knowledge. While previous works have only focused on detecting offensive memes, classifying them further into implicit and explicit categories depending on their severity is still a challenging and underexplored area. In this work, we present an end-to-end multitask model for addressing this challenge by empirically investigating two correlated tasks simultaneously: (i) offensive meme detection and (ii) explicit-implicit offensive meme detection by leveraging the two self-supervised pre-trained models. The first pre-trained model, referred to as the "knowledge encoder," incorporates contextual knowledge of the meme. On the other hand, the second model, referred to as the "fine-grained information encoder", is trained to understand the obscure psycho-linguistic information of the meme. Our proposed model utilizes contrastive learning to integrate these two pre-trained models, resulting in a more comprehensive understanding of the meme and its potential for offensiveness. To support our approach, we create a large-scale dataset, CM-Off-Meme, as there is no publicly available such dataset for the code-mixed Hindi-English (Hinglish) domain. Empirical evaluation, including both qualitative and quantitative analysis, on the CM-Off-Meme dataset demonstrates the effectiveness of the proposed model in terms of cross-domain generalization. The sample dataset and codes are available at this link: https://***/~ai-nlp-ml/*** as well as at our GitHub repository: https://***/Gitanjali1801/CM_***. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

A Reinforcement Learning Approach for Intelligent Conversational Chatbot for Enhancing Mental Health Therapy 2

A Reinforcement Learning Approach for Intelligent Conversati...

引用

2nd International Conference on Machine Learning and Data engineering, ICmlDE 2023

作者： Abubakar, Abdulqahar Mukhtar Gupta, Deepa Parida, Shantipriya Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Bengaluru India Silo AI Helsinki Finland

Mental health therapy plays a crucial role in addressing the growing challenges of mental health disorders. However, barriers such as limited accessibility and social stigma often hinder individuals from seeking timely and effective support. Advancement in the field of Artificial Intelligence (ai) and Natural Language Processing (NLP) has enabled the development of Chatbots which can simulate counseling conversations and provide support to those who seek help. However, existing chatbots often lack contextual understanding, limit user responses to multiple-choice questions, and face challenges in generating human-like responses. In this work, we present an Intelligent Conversational Agent using Reinforcement Learning with Human Feedback (RLHF) for mental health therapy support. The proposed system employs reinforcement learning to optimize its responses and adapts its conversational strategies to better understand user's responses, analyze the emotions expressed, and provide personalized therapeutic interventions to enhance Mental Health Therapy. © 2024 Elsevier B.V.. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Natural Language Processing-Driven Voice Recognition System for Enhancing Desktop Assistant Interactions

Natural Language Processing-Driven Voice Recognition System ...

引用

International Conference on Contemporary Computing and Informatics (IC3I)

作者： Swathi Gowroju Sandeep Kumar Shilpa Choudhary Department of Computer Science and Engineering (AI&ML) Sreyas Institute of Engineering and Technology Hyderabad Telangana India School of Computer Science and Artificial Intelligence SR University Warangal Andhra Pradesh India Department of Computer Science and Engineering (AI&ML) Neil Gogte Institute of Technology Hyderabad Telangana India

ISBN: (数字)9798350350067

ISBN: (纸本)9798350350074

A voice assistant is a software application designed to perform tasks and provide services based on the user's commands. This involves recognizing speech patterns and responding with synthetic speech. With these assistants, users can automate various tasks, including sending emails, managing schedules, and playing media. As technology progresses, people increasingly rely on these tools, particularly computers. Traditionally, commands are input through keyboards, but voice input offers a more accessible and user-friendly alternative. This is particularly beneficial for visually impaired individuals who find it challenging to use a keyboard. Voice assistants enhance the usability of computers by understanding and processing voice commands accurately. This allows for hands-free operation, improving convenience and productivity. Their advanced speech recognition technology involves sophisticated algorithms and machine learning to understand natural language and context. Once commands are processed, the assistant executes the tasks and provides feedback through voice or other means. For visually impaired users, voice assistants are transformative, enabling them to interact with their computers without needing a visual interface. They can perform tasks such as browsing the internet, sending messages, and controlling smart home devices through voice commands alone. Overall, voice assistants represent a significant advancement in making technology more accessible and user-friendly, allowing for a more inclusive environment where users can efficiently complete tasks through voice interaction.

关键词： computers Performance evaluation Visualization Technological innovation Personal voice assistants Keyboards Smart homes Machine learning Speech synthesis Usability

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：