检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Lee, Hyungyu Lee, Saehyung Jang, Hyemi Park, Junsung Bae, Ho Yoon, Sungroh Electrical and Computer Engineering Seoul National University Korea Republic of Department of Cyber Security Ewha Womans University Korea Republic of Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of

The disparity in accuracy between classes in standard training is amplified during adversarial training, a phenomenon termed the robust fairness problem. Existing methodologies aimed to enhance robust fairness by sacrificing the model's performance on easier classes in order to improve its performance on harder ones. However, we observe that under adversarial attacks, the majority of the model's predictions for samples from the worst class are biased towards classes similar to the worst class, rather than towards the easy classes. Through theoretical and empirical analysis, we demonstrate that robust fairness deteriorates as the distance between classes decreases. Motivated by these insights, we introduce the Distance-Aware Fair Adversarial training (DAFA) methodology, which addresses robust fairness by taking into account the similarities between classes. Specifically, our method assigns distinct loss weights and adversarial margins to each class and adjusts them to encourage a trade-off in robustness among similar classes. Experimental results across various datasets demonstrate that our method not only maintains average robust accuracy but also significantly improves the worst robust accuracy, indicating a marked improvement in robust fairness compared to existing methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

Global distribution of forest classes and leaf biomass for use as alternative foods to minimize malnutrition

World Food Policy

引用

World Food Policy 2021年第2期7卷 128-146页

作者： Fist, Tim Adesanya, Adewale A. Denkenberger, David Pearce, Joshua M. Alliance to Feed the Earth in Disasters (ALLFED) Environmental and Energy Program Social Science Department Michigan Technological University Department of Mechanical Engineering University of Alaska Fairbanks Department of Electrical and Computer Engineering Western University

Due to the ready availability of tree leaves in many geographies, the alternative food of leaf concentrate currently has the potential to alleviate hunger in over 800 million people. It is therefore potentially highly impactful to determine the edibility of leaf concentrates, which are in the same regions as the world's most undernourished populations. Unfortunately, the toxicity of leaf concentrate for most common tree leaf types has not been screened and the cost of doing so demands a prioritization. This preliminary study explores this potential solution to world hunger by finding the forest classes most likely to offer proximate access to the world's hungry, thus providing the basis for a prioritized list of leaf types to screen for toxicity. Specifically, this study describes a novel methodology for mapping available green leaf biomass and corresponding forest classes (e.g., tropical moist deciduous forest), and their spatial relationship to the global distribution of people who are underweight. These results will be useful for developing a targeted list of tree species to conduct leaf toxicity analysis on, in the interest of developing leaves as an alternative food source for both current malnutrition problems and global catastrophic scenarios. © 2021 Policy Studies Organization.

关键词： agriculture edible leaves forests GIS global catastrophic risk hunger

来源：评论

学校读者我要写书评

暂无评论

Q-Learning-Based Multi-Objective Global Path Planning for UGV Navigation*

Q-Learning-Based Multi-Objective Global Path Planning for UG...

引用

IEEE Latin American Robotics Symposium, LARS

作者： Hiago O. B. Batista Thayron M. Hudson Kevin B. de Carvalho Alexandre S. Brandão Department of Electrical Engineering Graduate Program in Computer Science Núcleo de Especialização em Robótica (NERo) Universidade Federal de Viçosa Viçosa - MG Brazil

ISBN: (数字)9798331508807

ISBN: (纸本)9798331508814

This paper presents an offline path planning strategy for unmanned ground vehicles (UGVs) using Q-learning. The proposed method addresses path optimization in warehouse-like environments, where tasks involve item pickup and delivery to specific locations. The Q-learning algorithm trains an agent to determine the most efficient routes, with validation conducted in an $8 \times 5$ meter workspace equipped with an Optitrack motion capture system. The workspace was discretized into a $16 \times 10$ grid, allowing the Q-learning to effectively navigate through complex obstacle-laden scenarios. Experimental results indicate that the Q-learning approach outperforms traditional methods such as Dijkstra, A-star, and Breadth-First Search in terms of path length, number of turns, planning time, and overall success rate; being up to 7 times faster to plan a path and reducing the number of bends by up to 41%. The Q-learning based paths feature more linear segments, which contribute to energy savings and improved navigational efficiency. Future work will explore applications in heterogeneous multi-agent systems and enhancements in training time and agent collaboration.

关键词： Training Meters Q-learning Navigation Path planning Motion capture Planning Robots Optimization Multi-agent systems

来源：评论

学校读者我要写书评

暂无评论

DanceOnStage: positioning training of dancing stage performance in virtual reality

DanceOnStage: positioning training of dancing stage performa...

引用

IEEE International Conference on Artificial Intelligence and Virtual Reality (AIVR)

作者： Ssu-Hsuan Wu Kai-Ting Wang Chia-Chen Chiang Hsin-Yu Chang Cheng-Yun Yeh Min-Chun Hu International Intercollegiate Ph.D.Program National Tsing Hua University Hsinchu Taiwan Master of Computer Science Program National Tsing Hua University Hsinchu Taiwan Interdisciplinary Program of Electrical Engineering and Computer Science National Tsing Hua University Hsinchu Taiwan Interdisciplinary Program of Technology and Art National Tsing Hua University Hsinchu Taiwan National Tsing Hua University Hsinchu Taiwan

This research aims to create a virtual reality-based practice system that simulates real-stage performance environments, assisting amateur dancers in overcoming stage fright and practicing group choreography that is difficult to achieve during individual practice. In this study, users can engage in performance simulations within the created system using the HTC VIVE virtual reality device. Within the system, we have developed a virtual stage space system that only requires a 2m x 3.5m physical area to simulate a large virtual stage experience. Users can also simulate choreographic positioning with virtual dance partners within this system, providing amateur dancers with group dance rehearsal practice opportunities that are not feasible during solo practice. Additionally, by creating three different audience modes, we aim to help users simulate various audience scenarios they may encounter during a performance. The system includes a recording feature that captures users' simulated performances, allowing them to review their actions and receive feedback after practice.

关键词：

来源：评论

学校读者我要写书评

暂无评论

UAV-Based Real-Time Air Quality Monitoring and Prediction Using Machine Learning Approach

引用

IEEE Internet of Things Journal 2025年

作者： Hsu, Huan-Chia Tarekegn, Getaneh Berie Getnet, Mastewal Endeshaw Yu, Kuo-Pin Tai, Li-Chia NYCU Department of Electrical and Computer Engineering Hsinchu30010 Taiwan NYCU Institute of Electrical and Control Engineering Hsinchu30010 Taiwan NYCU Institute of Environmental and Occupational Health Sciences Taipei112304 Taiwan NYCU Department of International Health Program Taipei112304 Taiwan

Globally, air pollution has become a major problem due to the growth of the human population and rapid urbanization, posing a serious threat to the environment and human health. Currently, most monitoring systems are based on fixed monitoring stations and are inefficient due to slow response time. An accurate and real-time air quality monitoring and prediction (AQMP) system is required to prevent health problems caused by air pollution. With the swift advancement of Internet of Things (IoT) technologies such as low-cost sensors and advanced wireless communication systems capable of sensing their environment, connecting with other systems, and interacting with users. To overcome the problems of existing systems, this paper aims to develop an unmanned aerial vehicle (UAV)-based real-time AQMP system by integrating artificial intelligence and data science technologies. The proposed UAV is equipped with lightweight air quality sensors including particulate matter (i.e., PM2.5 and PM10), carbon monoxide (CO), carbon dioxide (CO2), and ozone (O3), and meteorological sensors. The sensors are mounted on the UAV via the microcontroller, and the sensors reading data is transmitted to the web server via a message queueing telemetry transport (MQTT) protocol using a 5G mobile network module. Practical experiments were conducted in Hsinchu City and the industrial area of Yunlin County to collect real-time air pollutant data. Then, we employed a hybrid of attention mechanism, convolutional neural network (CNN) and gated recurrent unit (GRU) algorithms to predict future air quality levels. Experimental results show that the proposed system improves prediction performance compared with baseline time series algorithms. © 2014 IEEE.

关键词： Air quality

来源：评论

学校读者我要写书评

暂无评论

More Efficient Training Strategy to Leverage Neurons in Neural Network

More Efficient Training Strategy to Leverage Neurons in Neur...

引用

IEEE International Symposium on Industrial Electronics (ISIE)

作者： Cheng-Fu Liou Yi-Cheng Yu Research Labs (MZAIR Labs) Meta Zen Artificial Intelligence Tainan Taiwan Graduate Degree Program of College of Electrical and Computer Engineering National Yang Ming Chiao Tung University Hsinchu Taiwan

ISBN: (数字)9798350394085

ISBN: (纸本)9798350394092

Comparing the structures of current neural networks and the biological brain, it can be observed that many mechanisms correspond to each other according to their functions. From the perspective of the biological brain, different components have different long-term or short-term memory effects, such as the hippopotamus. However, current neural networks do not pay much attention to this aspect. In this work, we incorporated many hypotheses or theories inspired by the territory of neurobiology to retain the learned knowledge. To start with, we follow the inspiration of the synaptic homeostasis hypothesis (SHY) [1] and add an additional training stage to the training process of the model, which can ensure that only the most important information remains intact and the insignificant synapse can be pruned. In other words, we divide the overall training process into two learning stages: synaptogenesis and synaptic sparsifying. We experimentally demonstrate that our novel learning strategy significantly outperforms the traditional solution in training a sequence of tasks at different times on three public datasets, which supports that the proposed method is more efficient for resource-limited edge devices.

关键词： Training Industrial electronics Neurons Homeostasis Task analysis Biological neural networks Synapses

来源：评论

学校读者我要写书评

暂无评论

Platform for Evanescent Wave Sensor Fabrication Based on 3D-Printed Fiber Optic Holder

Platform for Evanescent Wave Sensor Fabrication Based on 3D-...

引用

Optical Fiber Sensors (LAWOFS), Latin American Workshop on

作者： Afonso Henrique Henschel Baptista Pedro Bertemes-Filho Aleksander Paterno Raquel Bettega Jose Pedro Mansueto Serbena Nathalia Prediger Ricardo Kamikawachi Department of Electrical Engineering Center for Science and Technology - Santa Catarina State University – CCT-UDESC Joinville Brazil Group of Organic Optoelectronic Devices (GOOD) Federal University of Parana – UFPR Curitiba Brazil Graduate Program in Electrical and Computer Engineering Federal University of Technology - Parana – UTFPR Curitiba Brazil

ISBN: (数字)9786589532026

ISBN: (纸本)9798350362725

This paper describes the low-cost manufacturing process of an evanescent wave fiber sensor platform that allows the etching of the fiber in hydrofluoric acid with the proposed 3D-printed fiber holder in an acid resistant material. The fiber holder is practical for allowing several processing procedures during its use and sensor fabrication, allowing sensor optical and electrical operation. In this work, some aspects of its characterization are presented in the optical domain showing the excitation of surface plasmons in the visible range, and also indication of its electrical operation of the thin-film as electrode.

关键词： Optical fiber polarization Optical fiber sensors Gold Optical polarization Biomedical optical imaging Optical device fabrication Optical films

来源：评论

学校读者我要写书评

暂无评论

SF(DA)2: SOURCE-FREE DOMAIN ADAPTATION THROUGH THE LENS OF DATA AUGMENTATION

arXiv

引用

arXiv 2024年

作者： Hwang, Uiwon Lee, Jonghyun Shin, Juhyeon Yoon, Sungroh Division of Digital Healthcare Yonsei University Korea Republic of Department of Electrical and Computer Engineering Seoul National University Korea Republic of Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of

In the face of the deep learning model’s vulnerability to domain shift, source-free domain adaptation (SFDA) methods have been proposed to adapt models to new, unseen target domains without requiring access to source domain data. Although the potential benefits of applying data augmentation to SFDA are attractive, several challenges arise such as the dependence on prior knowledge of class-preserving transformations and the increase in memory and computational requirements. In this paper, we propose Source-free Domain Adaptation Through the Lens of Data Augmentation (SF(DA)2), a novel approach that leverages the benefits of data augmentation without suffering from these challenges. We construct an augmentation graph in the feature space of the pretrained model using the neighbor relationships between target features and propose spectral neighborhood clustering to identify partitions in the prediction space. Furthermore, we propose implicit feature augmentation and feature disentanglement as regularization loss functions that effectively utilize class semantic information within the feature space. These regularizers simulate the inclusion of an unlimited number of augmented target features into the augmentation graph while minimizing computational and memory demands. Our method shows superior adaptation performance in SFDA scenarios, including 2D image and 3D point cloud datasets and a highly imbalanced dataset. Copyright © 2024, 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

Accumulating Curriculum Knowledge Based on Neural Network Isolated Algorithm

Accumulating Curriculum Knowledge Based on Neural Network Is...

引用

IEEE International Symposium on Industrial Electronics (ISIE)

作者： Cheng-Fu Liou Yi-Cheng Yu Meta Zen Artificial Intelligence Research Labs (MZAIR Labs) Tainan Taiwan Graduate Degree Program of College of Electrical and Computer Engineering National Yang Ming Chiao Tung University Hsinchu Taiwan

ISBN: (数字)9798350394085

ISBN: (纸本)9798350394092

Feeding the training data to neural networks in a specific sequence, from the simple data to the difficult data, utilizing curriculum learning can enhance performance improvements over the typical learning strategy based on shuffling training samples randomly, without any additional computational costs. This training approach has been successfully applied in all fields of machine learning. However, the current curriculum learning will gradually introduce difficult samples until the model eventually performs joint training. In addition, our experimental results find and demonstrate how the order of training tasks arranged according to the difficulty of different tasks influences the reuse percentage of model capacity, allowing us to observe the impact of curriculum learning on model performance from different perspectives. Finally, we also separately conduct three ablations analyses of our model to comprehensively enhance the understanding of the impacts of the proposed approach and further demonstrate its scalability to a standard large-scale dataset, i.e., the ImageNet dataset.

关键词： Training Industrial electronics Scalability Knowledge based systems Training data Machine learning Multitasking

来源：评论

学校读者我要写书评

暂无评论

On mitigating stability-plasticity dilemma in CLIP-guided image morphing via geodesic distillation loss

arXiv

引用

arXiv 2024年

作者： Oh, Yeongtak Lee, Saehyung Hwang, Uiwon Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Korea Republic of Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of Division of Digital Healthcare Yonsei University Korea Republic of

Large-scale language-vision pre-training models, such as CLIP, have achieved remarkable text-guided image morphing results by leveraging several unconditional generative models. However, existing CLIP-guided image morphing methods encounter difficulties when morphing photorealistic images. Specifically, existing guidance fails to provide detailed explanations of the morphing regions within the image, leading to misguidance. In this paper, we observed that such misguidance could be effectively mitigated by simply using a proper regularization loss. Our approach comprises two key components: 1) a geodesic cosine similarity loss that minimizes inter-modality features (i.e., image and text) on a projected subspace of CLIP space, and 2) a latent regularization loss that minimizes intra-modality features (i.e., image and image) on the image manifold. By replacing the naïve directional CLIP loss in a drop-in replacement manner, our method achieves superior morphing results on both images and videos for various benchmarks, including CLIP-inversion. © 2024, CC BY.

关键词： Distillation

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：