检索结果-内蒙古大学图书馆

Proceedings of the 51st International Conference on Parallel Processing

作者： Jaehee Jang Heoneok Ha Dahuin Jung Sungroh Yoon Department of Electrical and Computer Engineering Seoul National University Republic of Korea Department of Electrical and Computer Engineering and Interdisciplinary Program in Artificial Intelligence Seoul National University Republic of Korea and AIIS ASRI and INMC Seoul National University Republic of Korea

ISBN: (纸本)9781450397339

Personalized federated learning is aimed at allowing numerous clients to train personalized models while participating in collaborative training in a communication-efficient manner without exchanging private data. However, many personalized federated learning algorithms assume that clients have the same neural network architecture, and those for heterogeneous models remain understudied. In this study, we propose a novel personalized federated learning method called federated classifier averaging (FedClassAvg). Deep neural networks for supervised learning tasks consist of feature extractor and classifier layers. FedClassAvg aggregates classifier weights as an agreement on decision boundaries on feature spaces so that clients with not independently and identically distributed (non-iid) data can learn about scarce labels. In addition, local feature representation learning is applied to stabilize the decision boundaries and improve the local feature extraction capabilities for clients. While the existing methods require the collection of auxiliary data or model weights to generate a counterpart, FedClassAvg only requires clients to communicate with a couple of fully connected layers, which is highly communication-efficient. Moreover, FedClassAvg does not require extra optimization problems such as knowledge transfer, which requires intensive computation overhead. We evaluated FedClassAvg through extensive experiments and demonstrated it outperforms the current state-of-the-art algorithms on heterogeneous personalized federated learning tasks.

关键词： Communication Efficient Resource Constraint Representation Learning Neural Networks Model Heterogeneity Federated Learning

来源：评论

学校读者我要写书评

暂无评论

Approximate Thompson Sampling for Learning Linear Quadratic Regulators with O(√T) Regret

arXiv

引用

arXiv 2024年

作者： Kim, Yeoneung Kim, Gihun Yang, Insoon The Department of Applied Artificial Intelligence Seoul National University of Science and Technology Seoul01811 Korea Republic of The Department of Electrical and Computer Engineering ASRI Seoul National University Seoul08826 Korea Republic of

We propose an approximate Thompson sampling algorithm that learns linear quadratic regulators (LQR) with an improved Bayesian regret bound of O(√T). Our method leverages Langevin dynamics with a meticulously designed preconditioner as well as a simple excitation mechanism. We show that the excitation signal induces the minimum eigenvalue of the preconditioner to grow over time, thereby accelerating the approximate posterior sampling process. Moreover, we identify nontrivial concentration properties of the approximate posteriors generated by our algorithm. These properties enable us to bound the moments of the system state and attain an O(√T) regret bound without the unrealistic restrictive assumptions on parameter sets that are often used in the literature. Copyright © 2024, The Authors. All rights reserved.

关键词： Eigenvalues and eigenfunctions

来源：评论

学校读者我要写书评

暂无评论

DNN-Based Velocity Estimator Using Inertial Sensor for Robot Localization

引用

IFAC-PapersOnLine 2023年第2期56卷 5027-5032页

作者： Chul-hong Kim Dong-il “Dan” Cho Department of Electrical and Computer Engineering and Automation and Systems Research Institute (ASRI) Seoul National University Seoul Korea Inter-university Semiconductor Research Center (ISRC) Seoul National University Seoul Korea

This paper presents an inertial sensor-based localization method using a deep neural network (DNN)-based velocity estimator. Among various sensors commonly used in localization, inertial sensors are less affected by the surrounding environment. The inertial sensor consists of an accelerometer and a gyroscope, which can estimate the poses of a robot. However, inertial sensors need to be used with other sensors for localization because inherent large drift errors are difficult to prevent. To overcome this problem, a DNN-based velocity estimator that reduces the position error by learning the data pattern of the inertial sensor and by limiting the range of the estimated velocity is proposed. The DNN-based velocity estimator comprises a convolutional neural network, a fully connected layer, and a smoothing filter. To limit the range of the estimated velocity, the range of velocity is divided by the total number of classes, and the divided range is assigned to each class. The relationship between consecutive classes is learned using a smoothing filter. Dataset-based experiments are performed to train the DNN-based velocity estimator and to evaluate the performance of the proposed localization. The experimental result shows that the proposed localization reduces the position errors by 99% compared to integration-based localization and by 91% compared to extended Kalman filter-based localization.

关键词： Localization Mobile robot Inertial sensor Deep learning Kalman filter

来源：评论

学校读者我要写书评

暂无评论

Continual Learning in the Presence of Spurious Correlation

arXiv

引用

arXiv 2023年

作者： Lee, Donggyu Jung, Sangwon Moon, Taesup Department of Electrical and Computer Engineering Sungkyunkwan University Korea Republic of Department of Electrical and Computer Engineering Seoul National University Korea Republic of SNU-LG AI Research Center ASRI INMC IPAI/AIIS Seoul National University Korea Republic of

Most continual learning (CL) algorithms have focused on tackling the stability-plasticity dilemma, that is, the challenge of preventing the forgetting of previous tasks while learning new ones. However, they have overlooked the impact of the knowledge transfer when the dataset in a certain task is biased — namely, when some unintended spurious correlations of the tasks are learned from the biased dataset. In that case, how would they affect learning future tasks or the knowledge already learned from the past tasks? In this work, we carefully design systematic experiments using one synthetic and two real-world datasets to answer the question from our empirical findings. Specifically, we first show through two-task CL experiments that standard CL methods, which are unaware of dataset bias, can transfer biases from one task to another, both forward and backward, and this transfer is exacerbated depending on whether the CL methods focus on the stability or the plasticity. We then present that the bias transfer also exists and even accumulate in longer sequences of tasks. Finally, we propose a simple, yet strong plug-in method for debiasing-aware continual learning, dubbed as Group-class Balanced Greedy Sampling (BGS). As a result, we show that our BGS can always reduce the bias of a CL model, with a slight loss of CL performance at most. Copyright © 2023, The Authors. All rights reserved.

关键词： Plasticity

来源：评论

学校读者我要写书评

暂无评论

Mode Consensus Algorithms With Finite Convergence Time

arXiv

引用

arXiv 2024年

作者： Huang, Chao Shim, Hyungbo Yu, Siliang Anderson, Brian D.O. The College of Electronic and Information Engineering Tongji University Shanghai200092 China ASRI Department of Electrical and Computer Engineering Seoul National University Seoul Korea Republic of The School of Engineering Australian National University ActonACT2601 Australia

This paper studies the distributed mode consensus problem in a multi-agent system, in which the agents each possess a certain attribute and they aim to agree upon the mode (the most frequent attribute owned by the agents) via distributed computation. Three algorithms are proposed. The first one directly calculates the frequency of all attributes at every agent, with protocols based on blended dynamics, and then returns the most frequent attribute as the mode. Assuming knowledge at each agent of a lower bound of the mode frequency as a priori information, the second algorithm is able to reduce the number of frequencies to be computed at every agent if the lower bound is large. The third algorithm further eliminates the need for this information by introducing an adaptive updating mechanism. The algorithms find the mode in finite time, and estimates of convergence time are provided. The proposed first and second algorithms enjoy the plug-and-play property with a dwell time. © 2024, CC BY.

关键词： Multi agent systems

来源：评论

学校读者我要写书评

暂无评论

Data-Driven Inverse of Linear Systems and Application to Disturbance Observers

arXiv

引用

arXiv 2022年

作者： Eun, Yongsoon Lee, Jaeho Shim, Hyungbo The Department of Electrical Engineering and Computer Science DGIST Daegu42988 Korea Republic of ASRI Department of Electrical and Computer Engineering Seoul National University Seoul Korea Republic of

This work develops a data-based construction of inverse dynamics for LTI systems. Specifically, the problem addressed here is to find an input sequence from the corresponding output sequence based on pre-collected input and output data. The problem can be considered as a reverse of the recent use of behavioral approach, in which the output sequence is obtained for a given input sequence by solving an equation formed by pre-collected data. The condition under which the problem gives a solution is investigated and turns out to be L-delay invertibility of the plant and a certain degree of persistent excitation of the data input. The result is applied to form a data-driven disturbance observer. The plant dynamics augmented by the data-driven disturbance observer exhibits disturbance rejection without the model knowledge of the plant. © 2022, CC BY.

关键词： Disturbance rejection

来源：评论

学校读者我要写书评

暂无评论

Battling the Non-stationarity in Time Series Forecasting via Test-time Adaptation 39

Battling the Non-stationarity in Time Series Forecasting via...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Kim, HyunGi Kim, Siwon Mok, Jisoo 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 AIIS ASRI INMC Seoul National University Korea Republic of

ISBN: (纸本)157735897X

Deep Neural Networks have spearheaded remarkable advancements in time series forecasting (TSF), one of the major tasks in time series modeling. Nonetheless, the non-stationarity of time series undermines the reliability of pre-trained source time series forecasters in mission-critical deployment settings. In this study, we introduce a pioneering test-time adaptation framework tailored for TSF (TSF-TTA). TAFAS, the proposed approach to TSF-TTA, flexibly adapts source forecasters to continuously shifting test distributions while preserving the core semantic information learned during pre-training. The novel utilization of partially-observed ground truth and gated calibration module enables proactive, robust, and model-agnostic adaptation of source forecasters. Experiments on diverse benchmark datasets and cutting-edge architectures demonstrate the efficacy and generality of TAFAS, especially in long-term forecasting scenarios that suffer from significant distribution shifts. Copyright © 2025, Association for the Advancement of Artificial Intelligence.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Disentangled Motion Modeling for Video Frame Interpolation

arXiv

引用

arXiv 2024年

作者： Lew, Jaihyun Choi, Jooyoung Shin, Chaehun Jung, Dahuin Yoon, Sungroh Interdisciplinary Program in AI Seoul National University Korea Republic of Department of Electrical and Computer Engineering Seoul National University Korea Republic of School of Computer Science and Engineering Soongsil University Korea Republic of AIIS ASRI and INMC Seoul National University Korea Republic of

Video Frame Interpolation (VFI) aims to synthesize intermediate frames between existing frames to enhance visual smoothness and quality. Beyond the conventional methods based on the reconstruction loss, recent works have employed generative models for improved perceptual quality. However, they require complex training and large computational costs for pixel space modeling. In this paper, we introduce disentangled Motion Modeling (MoMo), a diffusion-based approach for VFI that enhances visual quality by focusing on intermediate motion modeling. We propose a disentangled two-stage training process. In the initial stage, frame synthesis and flow models are trained to generate accurate frames and flows optimal for synthesis. In the subsequent stage, we introduce a motion diffusion model, which incorporates our novel U-Net architecture specifically designed for optical flow, to generate bi-directional flows between frames. By learning the simpler low-frequency representation of motions, MoMo achieves superior perceptual quality with reduced computational demands compared to the generative modeling methods on the pixel space. MoMo surpasses state-of-the-art methods in perceptual metrics across various benchmarks, demonstrating its efficacy and efficiency in VFI. © 2024, CC BY-NC-SA.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Battling the Non-stationarity in Time Series Forecasting via Test-time Adaptation

arXiv

引用

arXiv 2025年

Deep Neural Networks have spearheaded remarkable advancements in time series forecasting (TSF), one of the major tasks in time series modeling. Nonetheless, the non-stationarity of time series undermines the reliability of pre-trained source time series forecasters in mission-critical deployment settings. In this study, we introduce a pioneering test-time adaptation framework tailored for TSF (TSF-TTA). TAFAS, the proposed approach to TSF-TTA, flexibly adapts source forecasters to continuously shifting test distributions while preserving the core semantic information learned during pre-training. The novel utilization of partially-observed ground truth and gated calibration module enables proactive, robust, and model-agnostic adaptation of source forecasters. Experiments on diverse benchmark datasets and cutting-edge architectures demonstrate the efficacy and generality of TAFAS, especially in long-term forecasting scenarios that suffer from significant distribution shifts. The code is available at https://***/kimanki/TAFAS. © 2025, CC BY-NC-SA.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Anti-Adversarially Manipulated Attributions for Weakly and Semi-Supervised Semantic Segmentation

Anti-Adversarially Manipulated Attributions for Weakly and S...

引用

2021 IEEE/CVF Conference on computer Vision and Pattern Recognition, CVPR 2021

作者： Lee, Jungbeom Kim, Eunji Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Seoul Korea Republic of ASRI INMC ISRC Institute of Engineering Research Seoul National University Korea Republic of

ISBN: (纸本)9781665445092

Weakly supervised semantic segmentation produces a pixel-level localization from a classifier, but it is likely to restrict its focus to a small discriminative region of the target object. AdvCAM is an attribution map of an image that is manipulated to increase the classification score. This manipulation is realized in an anti-adversarial manner, which perturbs the images along pixel gradients in the opposite direction from those used in an adversarial attack. It forces regions initially considered not to be discriminative to become involved in subsequent classifications, and produces attribution maps that successively identify more regions of the target object. In addition, we introduce a new regularization procedure that inhibits the incorrect attribution of regions unrelated to the target object and limits the attributions of the regions that already have high scores. On PASCAL VOC 2012 test images, we achieve mIoUs of 68.0 and 76.9 for weakly and semi-supervised semantic segmentation respectively, which represent a new state-of-the-art. The code is available at: https://***/jbeomlee93/AdvCAM. © 2021 IEEE

关键词： Semantic Segmentation

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：