检索结果-内蒙古大学图书馆

ENTROPY IS NOT ENOUGH FOR TEST-TIME ADAPTATION: FROM THE PERSPECTIVE OF DISENTANGLED FACTORS

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Lee, Jonghyun Jung, Dahuin Lee, Saehyung Park, Junsung Shin, Juhyeon Hwang, Uiwon Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Korea Republic of School of Computer Science and Engineering Soongsil University Korea Republic of Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of Division of Digital Healthcare Yonsei University Korea Republic of AIIS ASRI INMC and ISRC Seoul National University Korea Republic of

Test-time adaptation (TTA) fine-tunes pre-trained deep neural networks for unseen test data. The primary challenge of TTA is limited access to the entire test dataset during online updates, causing error accumulation. To mitigate it, TTA methods have utilized the model output’s entropy as a confidence metric that aims to determine which samples have a lower likelihood of causing error. Through experimental studies, however, we observed the unreliability of entropy as a confidence metric for TTA under biased scenarios and theoretically revealed that it stems from the neglect of the influence of latent disentangled factors of data on predictions. Building upon these findings, we introduce a novel TTA method named Destroy Your Object (DeYO), which leverages a newly proposed confidence metric named Pseudo-Label Probability Difference (PLPD). PLPD quantifies the influence of the shape of an object on prediction by measuring the difference between predictions before and after applying an object-destructive transformation. DeYO consists of sample selection and sample weighting, which employ entropy and PLPD concurrently. For robust adaptation, DeYO prioritizes samples that dominantly incorporate shape information when making predictions. Our extensive experiments demonstrate the consistent superiority of DeYO over baseline methods across various scenarios, including biased and wild. Project page is publicly available at https://***/DeYO/. Copyright © 2024, The Authors. All rights reserved.

关键词： Entropy

A Novel computerized cognitive Test for the Detection of Mild cognitive Impairment and Its Association with Neurodegeneration in Alzheimer’s Disease Prone Brain Regions

学校读者我要写书评

暂无评论

Advances in Alzheimer's Disease 2023年第3期12卷 38-54页

作者： Rosie E. Curiel Cid D. Diane Zheng Marcela Kitaigorodsky Malek Adjouadi Elizabeth A. Crocco Mike Georgiou Christian Gonzalez-Jimenez Alexandra Ortega Mohammed Goryawala Natalya Nagornaya Pradip Pattany Efrosyni Sfakianaki Ubbo Visser David A. Loewenstein Center for Cognitive Neuroscience and Aging and Department of Psychiatry and Behavioral Sciences Miller School of Medicine University of Miami Miami Florida USA Center for Advanced Technology and Education Department of Electrical and Computer Engineering College of Engineering and Computing Florida International University Miami Florida USA Department of Radiology and Nuclear Medicine Miller School of Medicine University of Miami Miami Florida USA Department of Computer Science University of Miami Miami Florida USA

During the prodromal stage of Alzheimer’s disease (AD), neurodegenerative changes can be identified by measuring volumetric loss in AD-prone brain regions on MRI. cognitive assessments that are sensitive enough to measure the early brain-behavior manifestations of AD and that correlate with biomarkers of neurodegeneration are needed to identify and monitor individuals at risk for dementia. Weak sensitivity to early cognitive change has been a major limitation of traditional cognitive assessments. In this study, we focused on expanding our previous work by determining whether a digitized cognitive stress test, the Loewenstein-Acevedo Scales for Semantic Interference and Learning, Brief computerized Version (LASSI-BC) could differentiate between cognitively Unimpaired (CU) and amnestic Mild cognitive Impairment (aMCI) groups. A second focus was to correlate LASSI-BC performance to volumetric reductions in AD-prone brain regions. Data was gathered from 111 older adults who were comprehensively evaluated and administered the LASSI-BC. Eighty-seven of these participants (51 CU;36 aMCI) underwent MR imaging. The volumes of 12 AD-prone brain regions were related to LASSI-BC and other memory tests correcting for False Discovery Rate (FDR). Results indicated that, even after adjusting for initial learning ability, the failure to recover from proactive semantic interference (frPSI) on the LASSI-BC differentiated between CU and aMCI groups. An optimal combination of frPSI and initial learning strength on the LASSI-BC yielded an area under the ROC curve of 0.876 (76.1% sensitivity, 82.7% specificity). Further, frPSI on the LASSI-BC was associated with volumetric reductions in the hippocampus, amygdala, inferior temporal lobes, precuneus, and posterior cingulate.

关键词： Mild cognitive Impairment Proactive Semantic Interference MRI Volume Cortical Thickness LASSI-L

Q-Learning based Local Path Planning for UAVs with Different Priorities

学校读者我要写书评

暂无评论

Q-Learning based Local Path Planning for UAVs with Different...

IEEE Latin American Robotics Symposium, LARS

作者： Kevin B. de Carvalho O. B. Hiago Batista Leonardo A. Fagundes-Júnior 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

Path planning is a crucial component of autonomous navigation and frequently demands different priorities such as path length, safety, or energy consumption, with the latter being particularly important in the context of unmanned autonomous vehicles. In many applications, the agent may have to react to environment shifting. Algorithms such as geometric and dynamic programming as well as techniques such as artificial potential fields have been employed to tackle this local planning problem. In recent years, machine learning has gained more evidence in many research fields due to its flexibility and generalization capabilities. In this study, we propose a Q-learning-based approach to local planning, which weighs three crucial factors- path length, safety, and energy consumption- that can be freely adjusted by the user to suit its application’s needs. The performance of the proposed method was tested in simulated static and dynamic scenarios as well as benchmarked with a baseline approach. The results show that it can perform well in both kinds of environments without struggling with the commom pitfalls of other local planning algorithms.

关键词：

Dynamic Fire and Smoke Detection Module with Enhanced Feature Integration and Attention Mechanisms

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Amjad, Ammar Tsung, Chang Hsien Tai, Li-Chia Department of Electrical and Computer Engineering National Yang Ming Chiao Tung University Hsinchu300 Taiwan Department of Computer Science and Information Engineering Chang Gung Universiry Taoyuan333 Taiwan Bachelor Program in Artificial Intelligence Chang Gung Universiry Taoyuan333 Taiwan Department of Physical Medicine and Rehabilitation at Linkou Taoyuan333 Taiwan

Essential components of early fire warning systems include effective fire and smoke detection mechanisms. Despite object detection-based technologies demonstrating promising results, the unique characteristics of fire and smoke, the complexity of fire environments, the scarcity of labeled datasets, and the presence of noise in images necessitate further improvements. We have developed a YOLOv7-tiny optimized architecture called the Dynamic Fire and Smoke Detection Module (DFDM) to address these challenges. Our approach includes an asymptotic feature pyramid network (AFPN) to bridge semantic gaps, thus enhancing fire detection capabilities in complex environments. A cross-level dual attention mechanism also refines critical smoke and fire detection features. To enhance performance while reducing model complexity, we introduced a novel partial convolution module designed to filter out extraneous information effectively. Experimental results demonstrate that DFDM provides robust fire and smoke detection capabilities, giving it a competitive advantage over more complex systems. Extensive visualization studies confirm DFDM's effectiveness in object recognition across varied settings. © 2024, The Authors. All rights reserved.

关键词： Smoke

Paper Quality Assessment based on Individual Wisdom Metrics from Open Peer Review

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zahorodnii, Andrii van den Bosch, Jasper J.F. Charest, Ian Summerfield, Christopher Fiete, Ila R. Department of Electrical Engineering and Computer Science MIT United States Department of Brain and Cognitive Sciences McGovern Institute MIT United States School of Psychology University of Leeds United Kingdom Département de Psychologie Universit é de Montréal Canada Department of Experimental Psychology University of Oxford United Kingdom

This study proposes a data-driven framework for enhancing the accuracy and efficiency of scientific peer review through an open, bottom-up process that estimates reviewer quality. Traditional closed peer review systems, while essential for quality control, are often slow, costly, and subject to biases that can impede scientific progress. Here, we introduce a method that evaluates individual reviewer reliability by quantifying agreement with community consensus scores and applying Bayesian weighting to refine paper quality assessments. We analyze open peer review data from two major scientific conferences, and demonstrate that reviewer-specific quality scores significantly improve the reliability of paper quality estimation. Perhaps surprisingly, we find that reviewer quality scores are unrelated to authorship quality. Our model incorporates incentive structures to recognize high-quality reviewers and encourage broader coverage of submitted papers, thereby mitigating the common "rich-get-richer" pitfall of social media. These findings suggest that open peer review, with mechanisms for estimating and incentivizing reviewer quality, offers a scalable and equitable alternative for scientific publishing, with potential to enhance the speed, fairness, and transparency of the peer review process. Copyright © 2025, The Authors. All rights reserved.

关键词： Data quality

Exploring Collaborative Distributed Diffusion-Based AI-Generated Content (AIGC) in Wireless Networks

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Du, Hongyang Zhang, Ruichen Niyato, Dusit Kang, Jiawen Xiong, Zehui Kim, Dong In Shen, Xuemin Poor, H. Vincent The School of Computer Science and Engineering The Energy Research Institute @ NTU Interdisciplinary Graduate Program Nanyang Technological University Singapore The School of Computer Science and Engineering Nanyang Technological University Singapore The School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China The School of Automation Guangdong University of Technology China The Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore The Department of Electrical and Computer Engineering Sungkyunkwan University Suwon16419 Korea Republic of The Department of Electrical and Computer Engineering University of Waterloo Canada The Department of Electrical and Computer Engineering Princeton University PrincetonNJ08544 United States

Driven by advances in generative artificial intelligence (AI) techniques and algorithms, the widespread adoption of AI-generated content (AIGC) has emerged, allowing for the generation of diverse and high-quality content. Especially, the diffusion model-based AIGC technique has been widely used to generate content in a variety of modalities. However, the real-world implementation of AIGC models, particularly on resource-constrained devices such as mobile phones, introduces significant challenges related to energy consumption and privacy concerns. To further promote the realization of ubiquitous AIGC services, we propose a novel collaborative distributed diffusion-based AIGC framework. By capitalizing on collaboration among devices in wireless networks, the proposed framework facilitates the efficient execution of AIGC tasks, optimizing edge computation resource utilization. Furthermore, we examine the practical implementation of the denoising steps on mobile phones, the impact of the proposed approach on the wireless network-aided AIGC landscape, and the future opportunities associated with its real-world integration. The contributions of this paper not only offer a promising solution to the existing limitations of AIGC services but also pave the way for future research in device collaboration, resource optimization, and the seamless delivery of AIGC services across various devices. Our code is available at https://***/HongyangDu/DistributedDiffusion. © 2023, CC BY.

关键词： Wireless networks

Towards Fully Decoupled End-to-End Person Search

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zhang, Pengcheng Zheng, Jin Bai, Xiao Ning, Xin School of Computer Science and Engineering State Key Laboratory of Software Development Environment Jiangxi Research Institute Beihang University Beijing China Institute of Semiconductors Chinese Academy of Sciences Cognitive Computing Technology Joint Laboratory Wave Group Beijing China

End-to-end person search aims to jointly detect and re-identify a target person in raw scene images with a unified model. The detection task unifies all persons while the re-id task discriminates different identities, resulting in conflict optimal objectives. Existing works proposed to decouple end-to-end person search to alleviate such conflict. Yet these methods are still sub-optimal on one or two of the sub-tasks due to their partially decoupled models, which limits the overall person search performance. In this paper, we propose to fully decouple person search towards optimal person search. A task-incremental person search network is proposed to incrementally construct an end-to-end model for the detection and re-id sub-task, which decouples the model architecture for the two sub-tasks. The proposed task-incremental network allows task-incremental training for the two conflicting tasks. This enables independent learning for different objectives thus fully decoupled the model for persons earch. Comprehensive experimental evaluations demonstrate the effectiveness of the proposed fully decoupled models for end-to-end person search. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

Language Model Pre-training on True Negatives

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhang, Zhuosheng Zhao, Hai Utiyama, Masao Sumita, Eiichiro Department of Computer Science and Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China Kyoto Japan

Discriminative pre-trained language models (PLMs) learn to predict original texts from intentionally corrupted ones. Taking the former text as positive and the latter as negative samples, the PLM can be trained effectively for contextualized representation. However, the training of such a type of PLMs highly relies on the quality of the automatically constructed samples. Existing PLMs simply treat all corrupted texts as equal negative without any examination, which actually lets the resulting model inevitably suffer from the false negative issue where training is carried out on pseudo-negative data and leads to less efficiency and less robustness in the resulting PLMs. In this work, on the basis of defining the false negative issue in discriminative PLMs that has been ignored for a long time, we design enhanced pre-training methods to counteract false negative predictions and encourage pre-training language models on true negatives by correcting the harmful gradient updates subject to false negative predictions. Experimental results on GLUE and SQuAD benchmarks show that our counter-false-negative pre-training methods indeed bring about better performance together with stronger robustness. Copyright © 2022, The Authors. All rights reserved.

关键词： Forecasting

Accelerating Real-Time Imaging for Radiotherapy: Leveraging Multi-GPU Training with PyTorch 22

学校读者我要写书评

暂无评论

Accelerating Real-Time Imaging for Radiotherapy: Leveraging ...

22nd IEEE International Conference on Machine Learning and Applications, ICMLA 2023

作者： Obe, Ruth Kaufmann, Brandt Baird, Kaelen Kadel, Sam Soltani, Yasmin Cham, Mostafa Gobbert, Matthias K. Barajas, Carlos A. Jiang, Zhuoran Sharma, Vijay R. Ren, Lei Peterson, Stephen W. Poif, Jerimy C. U. of Houston-Clear Lake Dept. of Computer Science and of Software Engineering United States U. of San Francisco Dept. of Mathematics and Statistics United States Skidmore College Dept. of Computer Science and of Mathematics United States Mount Holyoke College Dept. of Computer Science and of Psychology United States U. of Houston Dept. of Biomedical Engineering United States U. of Maryland Dept. of Information Systems Baltimore County United States U. of Maryland Dept. of Mathematics and Statistics Baltimore County United States Medical Physics Graduate Program Duke University United States U. of Maryland School of Medicine Dept. of Radiation Oncology United States U. of Maryland School of Medicine Department of Radiation Oncology United States U. of Cape Town Dept. of Physics South Africa H3D Inc. United States

ISBN: (纸本)9798350345346

Proton beam therapy is an advanced form of cancer radiotherapy that uses high-energy proton beams to deliver precise and targeted radiation to tumors. This helps to mit-igate unnecessary radiation exposure in healthy tissues. Real-time imaging of prompt gamma rays with Compton cameras has been suggested to improve therapy efficacy. However, the camera's non-zero time resolution leads to incorrect interaction classifications and noisy images that are insufficient for accurately assessing proton delivery in patients. To address the challenges posed by the Compton camera's image quality, machine learning techniques are employed to classify and refine the generated data. These machine-learning techniques include recurrent and feedforward neural networks. A PyTorch model was designed to improve the data captured by the Compton camera. This decision was driven by PyTorch's flexibility, powerful capabilities in handling sequential data, and enhanced G PU usage. This accelerates the model's computations on large-scale radiotherapy data. Through hyperparameter tuning, the validation accuracy of our PyTorch model has been improved from an initial 7 % to over 60 %. Moreover, the PyTorch Distributed Data Parallelism strategy was used to train the RNN models on multiple G PU s, which significantly reduced the training time with a minor impact on model accuracy. © 2023 IEEE.

关键词： Clas-sification Compton camera Distributed Data Parallelism Proton beam therapy PyTorch Recurrent neural network