检索结果-内蒙古大学图书馆

Blaming Humans and Machines: What Shapes People’s Reactions to Algorithmic Harm

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lima, Gabriel Grgić-Hlača, Nina Cha, Meeyoung School of Computing KAIST & Data Science Group IBS Korea Republic of Max Planck Institute for Software Systems Max Planck Institute for Research on Collective Goods Germany Data Science Group IBS & School of Computing KAIST Korea Republic of

Artificial intelligence (AI) systems can cause harm to people. This research examines how individuals react to such harm through the lens of blame. Building upon research suggesting that people blame AI systems, we investigated how several factors influence people’s reactive attitudes towards machines, designers, and users. The results of three studies (N = 1,153) indicate differences in how blame is attributed to these actors. Whether AI systems were explainable did not impact blame directed at them, their developers, and their users. Considerations about fairness and harmfulness increased blame towards designers and users but had little to no effect on judgments of AI systems. Instead, what determined people’s reactive attitudes towards machines was whether people thought blaming them would be a suitable response to algorithmic harm. We discuss implications, such as how future decisions about including AI systems in the social and moral spheres will shape laypeople’s reactions to AI-caused harm. Copyright © 2023, The Authors. All rights reserved.

关键词： Decision making

DRL-Based Time-Varying Workload Scheduling With Priority and Resource Awareness

学校读者我要写书评

暂无评论

IEEE Transactions on Network and Service Management 2025年第3期22卷 2838-2852页

作者： Liu, Qifeng Fan, Qilin Zhang, Xu Li, Xiuhua Wang, Kai Xiong, Qingyu Chongqing University School of Big Data and Software Engineering Chongqing400044 China Chongqing University Key Laboratory of Dependable Service Computing in Cyber Physical Society of Ministry of Education Chongqing400044 China Nanjing University School of Electronic Science and Engineering Nanjing210023 China Haihe Laboratory of Information Technology Application Innovation Tianjin300072 China Harbin Institute of Technology School of Computer Science and Technology Weihai264209 China Shandong Key Laboratory of Industrial Network Security Weihai264209 China

With the proliferation of cloud services and the continuous growth in enterprises' demand for dynamic multi-dimensional resources, the implementation of effective strategy for time-varying workload scheduling has become increasingly significant. In this paper, we propose a deep reinforcement learning (DRL)-based method for time-varying workload scheduling, aiming to allocate resources efficiently across servers in the cluster. Specifically, we integrate a classifier and queue scorer to construct a priority queue that exploits temporal resource utilization patterns across different workload classes. Then, we design parallel graph attention layers to capture the dimensional features and temporal dynamics of cloud server cluster. Moreover, we propose a DRL algorithm to generate scheduling strategies that can adapt to dynamic environments. Validation on real-world traces from Google cluster demonstrates that our method outperforms existing approaches in key metrics of cloud server cluster management. © 2004-2012 IEEE.

关键词： Cloud platforms

Gesture Showdown: Rock Paper Scissors with AI Vision

学校读者我要写书评

暂无评论

Gesture Showdown: Rock Paper Scissors with AI Vision

Computational, Communication and Information Technology (ICCCIT), International Conference on

作者： Hemal Babu H Yeshwanth. C. D Shravan. N K. Regin Bose Department of Artificial Intelligence and Data science Rajalakshmi Institute of Technology Chennai India Department of AI&DS and Cloud Computing Rajalakshmi Institute of Technology Chennai India

ISBN: (数字)9798331512965

ISBN: (纸本)9798331512972

This paper presents a gesture-based RockPaper-Scissors game that leverages computer vision and hand gesture recognition to enable intuitive human-computer interaction. Utilizing MediaPipe's pre-trained hand tracking module, the system detects and identifies 21 hand landmarks in real-time to classify gestures as rock, paper or scissors based on angular relationships between key points. OpenCV is used for webcam integration and video frame preprocessing, ensuring seamless interaction. The game compares the recognized gesture with a computer-generated move, dynamically displaying the results and resetting after each round for continuous play. The system identifies gestures through precise angle calculations, differentiating between closed fists, open palms and extended fingers. Sample outputs demonstrate real-time gameplay scenarios, highlighting gesture recognition accuracy and user engagement. Future enhancements include refining recognition accuracy, adding multiplayer functionality, introducing customizable gestures, and deploying the game on web and mobile platforms for broader accessibility. This project showcases the potential of integrating AI and computer vision in creating engaging and user-friendly applications.

关键词： Hands Human computer interaction Computer vision Accuracy Webcams Tracking Games Gesture recognition Rocks Real-time systems

Text to Point Cloud Localization with Relation-Enhanced Transformer

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Guangzhi Fan, Hehe Kankanhalli, Mohan Institute of Data Science National University of Singapore Singapore School of Computing National University of Singapore Singapore

Automatically localizing a position based on a few natural language instructions is essential for future robots to communicate and collaborate with humans. To approach this goal, we focus on the text-to-point-cloud cross-modal localization problem. Given a textual query, it aims to identify the described location from city-scale point clouds. The task involves two challenges. 1) In city-scale point clouds, similar ambient instances may exist in several locations. Searching each location in a huge point cloud with only instances as guidance may lead to less discriminative signals and incorrect results. 2) In textual descriptions, the hints are provided separately. In this case, the relations among those hints are not explicitly described, leading to the difficulties of learning relations. To overcome these two challenges, we propose a unified Relation-Enhanced Transformer (RET) to improve representation discriminability for both point cloud and natural language queries. The core of the proposed RET is a novel Relation-enhanced Self-Attention (RSA) mechanism, which explicitly encodes instance (hint)-wise relations for the two modalities. Moreover, we propose a fine-grained cross-modal matching method to further refine the location predictions in a subsequent instance-hint matching stage. Experimental results on the KITTI360Pose dataset demonstrate that our approach surpasses the previous state-of-the-art method by large margins. © 2023, CC BY.

关键词： Location

data-driven sleep structure deciphering based on cardiorespiratory signals

学校读者我要写书评

暂无评论

Computer Methods and Programs in Biomedicine 2025年 267卷 108769-108769页

作者： Huang, Ming Iwata, Osuke Yokoyama, Kiyoko Tamura, Toshiyo Institute of Advanced Computing and Digital Engineering Shenzhen Institute of Advanced Technology Shenzhen China of Science and Technology Nara Institute of Science and Technology Ikoma Japan of Medical Sciences Nagoya City University Japan School of Data Science Nagoya City University Japan Institute for Healthcare Robotics Waseda University Japan

Background and Objective: : Cardiorespiratory signals provide a novel perspective for understanding sleep structure through the physiological mechanism of cardiopulmonary coupling. This mechanism divides the coupling spectrum into high-frequency (HF) and low-frequency (LF) bands, indicating that signal segments of 4–8 min are optimal for analysis. However, the lack of labels tailored to these signals has led to reliance on the American Academy of Sleep Medicine (AASM) definitions, which are primarily designed for electroencephalogram (EEG) and electrooculogram (EOG) data. This study aims to address the challenge of transitioning from AASM-defined labels to cardiorespiratory-oriented ones and to evaluate the feasibility of using these signals for accurate sleep structure recognition. Methods: To align with the physiological characteristics of cardiorespiratory signals, AASM labels were modified by excluding the N2 stage due to its overlap of stable and unstable non-rapid eye movement (NREM) phases, which introduces ambiguity. The modified dataset focused on the wake, N1, deep sleep (N3), and rapid eye movement (REM) stages. A physiologically-inspired deep-learning model (PIDM) was developed to extract features from cardiorespiratory time series and classify sleep stages. Post-analysis assessed the physiological validity of the model's N2 predictions by evaluating the HF-to-LF ratio and respiratory variability. Results: The pipeline, combining the modified labeling scheme with the PIDM model, achieved balanced accuracy scores of 0.83, 0.86, and 0.78 for wake, deep sleep, and REM stages, respectively in the normal group;and 0.92, 0.95, and 0.90 in the mild and moderate sleep apnea groups. Post-analysis revealed that most N2 samples were attributed to stable NREM sleep, characterized by higher HF-to-LF ratios and lower respiratory variability, aligning with physiological understanding. Conclusions: This study highlights the physiological relevance of cardiorespiratory s

关键词： Electroencephalography

SAC-ViT: Semantic-Aware Clustering Vision Transformer with Early Exit

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hu, Youbing Cheng, Yun Lu, Anqi Wei, Dawei Li, Zhijun Faculty of Computing Harbin Institute of Technology China Swiss Data Science Center Zurich Switzerland School of Cyber Engineering Xidian University China

The Vision Transformer (ViT) excels in global modeling but faces deployment challenges on resource-constrained devices due to the quadratic computational complexity of its attention mechanism. To address this, we propose the Semantic-Aware Clustering Vision Transformer (SAC-ViT), a non-iterative approach to enhance ViT’s computational efficiency. SAC-ViT operates in two stages: Early Exit (EE) and Semantic-Aware Clustering (SAC). In the EE stage, downsampled input images are processed to extract global semantic information and generate initial inference results. If these results do not meet the EE termination criteria, the information is clustered into target and non-target tokens. In the SAC stage, target tokens are mapped back to the original image, cropped, and embedded. These target tokens are then combined with reused non-target tokens from the EE stage, and the attention mechanism is applied within each cluster. This two-stage design, with end-to-end optimization, reduces spatial redundancy and enhances computational efficiency, significantly boosting overall ViT performance. Extensive experiments demonstrate the efficacy of SAC-ViT, reducing 62% of the FLOPs of DeiT and achieving 1.98× throughput without compromising performance. Copyright © 2025, The Authors. All rights reserved.

关键词： Latent semantic analysis

FoCTTA: Low-Memory Continual Test-Time Adaptation with Focus

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hu, Youbing Cheng, Yun Zhou, Zimu Lu, Anqi Cao, Zhiqiang Li, Zhijun Faculty of Computing Harbin Institute of Technology China Swiss Data Science Center Zurich Switzerland City University of Hong Kong Hong Kong

Continual adaptation to domain shifts at test time (CTTA) is crucial for enhancing the intelligence of deep learning enabled IoT applications. However, prevailing TTA methods, which typically update all batch normalization (BN) layers, exhibit two memory inefficiencies. First, the reliance on BN layers for adaptation necessitates large batch sizes, leading to high memory usage. Second, updating all BN layers requires storing the activations of all BN layers for backpropagation, exacerbating the memory demand. Both factors lead to substantial memory costs, making existing solutions impractical for IoT devices. In this paper, we present FoCTTA, a low-memory CTTA strategy. The key is to automatically identify and adapt a few drift-sensitive representation layers, rather than blindly update all BN layers. The shift from BN to representation layers eliminates the need for large batch sizes. Also, by updating adaptation-critical layers only, FoCTTA avoids storing excessive activations. This focused adaptation approach ensures that FoCTTA is not only memory-efficient but also maintains effective adaptation. Evaluations show that FoCTTA improves the adaptation accuracy over the state-of-the-arts by 4.5%, 4.9%, and 14.8% on CIFAR10-C, CIFAR100-C, and ImageNet-C under the same memory constraints. Across various batch sizes, FoCTTA reduces the memory usage by 3-fold on average, while improving the accuracy by 8.1%, 3.6%, and 0.2%, respectively, on the three datasets. Copyright © 2025, The Authors. All rights reserved.

关键词：

Fast Searching The Densest Subgraph And Decomposition With Local Optimality

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zhu, Yugao Liu, Shenghua Feng, Wenjie Cheng, Xueqi Institute of Computing Technology Chinese Academic of Sciences Beijing China Institute of Data Science National University of Singapore Singapore

Densest Subgraph Problem (DSP) is an important primitive problem with a wide range of applications, including fraud detection, community detection, and DNA motif discovery. Edge-based density is one of the most common metrics in DSP. Although a maximum flow algorithm can exactly solve it in polynomial time, the increasing amount of data and the high complexity of algorithms motivate scientists to find approximation algorithms. Among these, its duality of linear programming derives several iterative algorithms including Greedy++, Frank-Wolfe, and FISTA which redistribute edge weights to find the densest subgraph, however, these iterative algorithms vibrate around the optimal solution, which is not satisfactory for fast convergence. We propose our main algorithm Locally Optimal Weight Distribution (LOWD) to distribute the remaining edge weights in a locally optimal operation to converge to the optimal solution monotonically. Theoretically, we show that it will reach the optimal state of specific quadratic programming, which is called locally-dense decomposition. Besides, we show that it is not necessary to consider most of the edges in the original graph. Therefore, we develop a pruning algorithm using a modified Counting Sort to prune graphs by removing unnecessary edges and nodes, and then we can search the densest subgraph in a much smaller graph. Copyright © 2023, The Authors. All rights reserved.

关键词： Quadratic programming

A learning-based approach for efficient visualization construction

学校读者我要写书评

暂无评论

Visual Informatics 2022年第1期6卷 14-25页

作者： Yongjian Sun Jie Li Siming Chen Gennady Andrienko Natalia Andrienko Kang Zhang College of Intelligence and Computing Tianjin UniversityChina School of Data Science Fudan UniversityChina Fraunhofer Institute IAIS Germany City University London UK Beijing Normal University-Hong Kong Baptist University United International College China

We propose an approach to underpin interactive visual exploration of large data volumes by training Learned Visualization Index(LVI).Knowing in advance the data,the aggregation functions that are used for visualization,the visual encoding,and available interactive operations for data selection,LVI allows to avoid time-consuming data retrieval and processing of raw data in response to user’s ***,LVI directly predicts aggregates of interest for the user’s data *** demonstrate the efficiency of the proposed approach in application to two use cases of spatio-temporal data at different scales.

关键词： Learned index Neural network Visualization index Interactive exploration Spatiotemporal visualization