检索结果-内蒙古大学图书馆

SCSQ-MDD: A Detection Approach for Multi-Style Strokes of Chinese Characters

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Zhou, Tian Xie, Wu Zhang, Huimin Fan, Yong Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin University of Electronic Technology Guilin China School of Computer Science and Information Security Guilin University of Electronic Technology Guilin China Key Lab of Education Blockchain and Intelligent Technology Ministry of Education Guangxi Normal University Guilin541004 China Guangxi Key Lab of Multi-Source Information Mining and Security Guangxi Normal University Guilin541004 China School of Mechanical and Electrical Engineering Guilin University of Electronic Science and Technology Guilin China

In the Chinese character writing task of the robotic arms, the stroke category and position information should be extracted by object detection. The detection algorithms based on predefined anchor frames have difficulty in resolving the differences among many different styles of Chinese character strokes. While the deformable detection transformer (deformable DETR) algorithms without predefined anchor frames result in some invalid sampling points having no contribution to the feature update of the current reference point due to the random sampling of sampling points in the deformable attention module. These processes cause the effectiveness of correlation calculations between reference points in Chinese strokes and their surrounding sampled points is limited. So that the speed of vector learning stroke features in the detection head is reduced. In view of this problem, a new detection method of multi-style strokes of Chinese characters via SCSQ-MDD (Simple Conditional Spatial Query Mask Deformable DETR) is proposed in this paper. Firstly, a mask prediction layer is jointly determined using the shallow feature map of the Chinese character image and the query vector of the transformer encoder, which is used to filter the points with actual contribution and resample the points without contribution, so that the randomness of correlation calculation among reference points is solved. Secondly, by separating the content query and spatial query of the transformer deocder, the content embedding and spatial embedding can be separately focused on when cross-attention computations are performed. Thus the dependence of the prediction task on the content embedding is relaxed and the training process is simplified. Finally, the detection model without predefined anchor frames based on deformable DETR called SCSQ-MDD is constructed using the mask mechanism and the simple conditional spatial query mechanism, and trained and validated on a multi-style Chinese character stroke dataset

关键词： Object detection

A Born-Again Takagi-Sugeno-Kang Fuzzy Classifier with Decoupled Fuzzy Dark Knowledge Distillation

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Zhang, Xiongtao Yin, Zezong Jiang, Yunliang Jiang, Yizhang Sun, Danfeng Liu, Yong School of Information Engineering Huzhou University Huzhou31300 China Zhejiang Key Laboratory of Intelligent Education Technology and Application Zhejiang Normal University Jinhua321004 China School of Computer Science and Technology Zhejiang Normal University Jinhua321004 China School of Artificial Intelligence and Computer Science Jiangnan University Wuxi 214122 China Key Laboratory of Image Processing and Intelligent Control Huazhong University of Science and Technology Ministry of Education Wuhan430074 China School of Computer Science Hangzhou Dianzi University Hangzhou310018 China College of Control Science and Engineering Zhejiang University Hangzhou310027 China

In order to enpower the hight performance as well as interpretability of low-order TSK fuzzy classifier, a born-again TSK fuzzy classier embedded with decoupled fuzzy dark knowledge distillation called HTSK-LLM-DKD is proposed in this study. HTSK-LLM DKD achieves the following distinctive characteristics: 1) It takes high-order TSK as teacher model and low-order TSK fuzzy classifier as student model, and leverages the proposed Least Learning Machine based Decoupling Knowledge Distillation to distill the fuzzy dark knowledge from high-order TSK fuzzy classifier to low-order TSK fuzzy classifier, which resulting in low-order TSK fuzzy classifier endowed with enhanced performance surpassing or at least comparable to high-order TSK fuzzy classifier;specifically 2) The Negative Euclidean distance between the output of teacher model and each class is employed to obtain the teacher logits, and then it computes teacher/student soft labels by the softmax function with distilling temperature parameter;3) By reformulating the Kullback Leibler divergence, it decouples fuzzy dark knowledge into target class knowledge and non-target class knowledge, and transfers them to student model, which taking HTSK-LLM-DKD with outstanding performance and relatively fast running speed. Extensive experiment results on UCI benchmarking datasets and real-world dataset Cleveland Heart Disease demonstrate the effectiveness of the proposed HTSK-LLM-DKD. © 2024, The Authors. All rights reserved.

关键词： Students

Joint Observer Gain and Input Design for Asymptotic Active Fault Diagnosis

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Xu, Feng Wan, Yiming Wang, Ye Puig, Vicenç Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen518055 China School of Artificial Intelligence and Automation Huazhong University of Science and Technology Key Laboratory of Image Processing and Intelligent Control Ministry of Education China Wuhan430074 China Department of Electrical and Electronic Engineering The University of Melbourne ParkvilleVIC3010 Australia Universitat Politècnica de Catalunya Barcelona08028 Spain

This paper proposes a joint gain and input design method for observer-based asymptotic active fault diagnosis, which is based on a newly-defined notion named the excluding degree of the origin from a zonotope. Using the excluding degree, a quantitative specification is obtained to characterize the performance of set-based robust fault diagnosis. Furthermore, a single gain design method and a joint gain and input design method are proposed, respectively. This is the first work to achieve a joint observer gain and input design for set-based active fault diagnosis. Compared with the existing methods that design gains and input separately, the proposed joint gain and input design method has advantages to exploit the fault diagnosis potential of observer-based schemes. Finally, several examples are used to illustrate the effectiveness of the proposed methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Design

Weakly supervised person search with region siamese networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Han, Chuchu Su, Kai Yu, Dongdong Yuan, Zehuan Gao, Changxin Sang, Nong Yang, Yi Wang, Changhu Key Laboratory of Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology ByteDance University of Technology Sydney

Supervised learning is dominant in person search, but it requires elaborate labeling of bounding boxes and identities. Large-scale labeled training data is often difficult to collect, especially for person identities. A natural question is whether a good person search model can be trained without the need of identity supervision. In this paper, we present a weakly supervised setting where only bounding box annotations are available. Based on this new setting, we provide an effective baseline model termed Region Siamese Networks (R-SiamNets). Towards learning useful representations for recognition in the absence of identity labels, we supervise the R-SiamNet with instance-level consistency loss and cluster-level contrastive loss. For instance-level consistency learning, the R-SiamNet is constrained to extract consistent features from each person region with or without out-of-region context. For cluster-level contrastive learning, we enforce the aggregation of closest instances and the separation of dissimilar ones in feature space. Extensive experiments validate the utility of our weakly supervised method. Our model achieves the rank-1 of 87.1% and mAP of 86.0% on CUHK-SYSU benchmark, which surpasses several fully supervised methods, such as OIM [36] and MGTS [4], by a clear margin. More promising performance can be reached by incorporating extra training data. We hope this work could encourage the future research in this field. Copyright © 2021, The Authors. All rights reserved.

关键词： Supervised learning

Computationally Expensive High-Dimensional Multiobjective Optimization via Surrogate-Assisted Reformulation and Decomposition

学校读者我要写书评

暂无评论

IEEE Transactions on Evolutionary Computation 2024年 1-1页

作者： Pan, Linqiang Lin, Jianqing Wang, Handing He, Cheng Tan, Kay Chen Jin, Yaochu School of Artificial Intelligence and Automation Key Laboratory of Image Information Processing and Intelligent Control of Education Ministry of China Huazhong University of Science and Technology Wuhan China School of Artificial Intelligence Xidian University Xi’an China School of Electrical and Electronic Engineering Huazhong University of Science and Technology Wuhan China Department of Computing The Hong Kong Polytechnic University Hong Kong Hong Kong School of Engineering Trustworth and General Artificial Intelligence Laboratory Westlake University Hangzhou China

In recent decades, various surrogate-assisted evolutionary algorithms (SAEAs) have been proposed to solve computationally expensive multiobjective optimization problems (EMOPs). Nevertheless, designing an SAEA to handle high-dimensional EMOPs and balance convergence, diversity, and computational complexity remains challenging. Here, we propose a two-phase SAEA (TP-SAEA), which follows the idea of convergence first and diversity second, for solving high-dimensional EMOPs. In Phase I, a surrogate-assisted problem reformulation method is proposed to fast-track the Pareto optimal set in association with some reference solutions. Specifically, the high-dimensional EMOP is reformulated into an expensive single-objective one with low-dimensional decision space. Then, the surrogate-assisted optimization is utilized to obtain well-converged solutions. In Phase II, the high-dimensional EMOP is decomposed into two subproblems to explore subregions of the decision space that can effectively promote the diversity of the solutions. The two subproblems are optimized independently via surrogate-assisted optimization, aiming to push the population towards different regions of the Pareto optimal front. Experiments are conducted on EMOPs with 100 to 500 decision variables compared with four state-of-the-art SAEAs. The proposed TP-SAEA obtains well-converged and diverse solutions with only 509 real function evaluations. Moreover, its superiority is examined in six real-world instances with up to 12,000 decision variables. IEEE

关键词： Iron

Weakly Supervised Person Search with Region Siamese Networks

学校读者我要写书评

暂无评论

Weakly Supervised Person Search with Region Siamese Networks

International Conference on Computer Vision (ICCV)

作者： Chuchu Han Kai Su Dongdong Yu Zehuan Yuan Changxin Gao Nong Sang Yi Yang Changhu Wang Key Laboratory of Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology ByteDance University of Technology Sydney

ISBN: (纸本)9781665428132

关键词： Training Computer vision Annotations Supervised learning Training data Benchmark testing Feature extraction

Multi-Centroid Representation Network for Domain Adaptive Person Re-ID

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wu, Yuhang Huang, Tengteng Yao, Haotian Zhang, Chi Shao, Yuanjie Han, Chuchu Gao, Changxin Sang, Nong Key Laboratory of Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology China Megvii Technology China

Recently, many approaches tackle the Unsupervised Domain Adaptive person re-identification (UDA re-ID) problem through pseudo-label-based contrastive learning. During training, a uni-centroid representation is obtained by simply averaging all the instance features from a cluster with the same pseudo label. However, a cluster may contain images with different identities (label noises) due to the imperfect clustering results, which makes the uni-centroid representation inappropriate. In this paper, we present a novel Multi-Centroid Memory (MCM) to adaptively capture different identity information within the cluster. MCM can effectively alleviate the issue of label noises by selecting proper positive/negative centroids for the query image. Moreover, we further propose two strategies to improve the contrastive learning process. First, we present a Domain-Specific Contrastive Learning (DSCL) mechanism to fully explore intra-domain information by comparing samples only from the same domain. Second, we propose Second-Order Nearest Interpolation (SONI) to obtain abundant and informative negative samples. We integrate MCM, DSCL, and SONI into a unified framework named Multi-Centroid Representation Network (MCRN). Extensive experiments demonstrate the superiority of MCRN over state-of-the-art approaches on multiple UDA re-ID tasks and fully unsupervised re-ID tasks. Copyright © 2021, The Authors. All rights reserved.

关键词： Machine learning

Closer to pre-trained network transfer better 9

学校读者我要写书评

暂无评论

Closer to pre-trained network transfer better

9th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2020

作者： Chen, Siyu Li, Wei Huazhong University of Science and Technology School of Artificial Intelligence and Automation Wuhan430074 China Image Processing and Intelligent Control Key Laboratory Education Ministry of China Wuhan430074 China

In recent years, Deep Neural Network (DNN) has been widely used in the domain of computer vision, but its further development is restricted because of the lack of train samples. Fine-tuning is one of deep transfer learning methods to solve this problem. Nowdays, some existing Fine-tuning approaches assume that if the fine-tuned network could be as close to the pre-trained network as possible, the source knowledge would be fully utilized to improve the study effect. Although these approaches work pretty well, the assumptions on which they are based have not been systematically proved. Due to the complexity of DNNs, offering a reasonable and accurate explanation of the performance of the network is significant for designing new algorithms. In this paper, we design a series of experiments with or without transfer learning respectively, to characterize the relationship between network differences and network performance. The results show that there is an obvious negative correlation between them under certain circumstances. This conclusion can not only provide a more reasonable explanation for the relevant transfer learning algorithm, but also provide support and in-spiration for the future work. © 2020 IEEE.

关键词： Deep neural networks

T-TIME: Test-Time Information Maximization Ensemble for Plug-and-Play BCIs

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Siyang Wang, Ziwei Luo, Hanbin Ding, Lieyun Wu, Dongrui Ministry of Education Key Laboratory of Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen China Henan Key Laboratory of Brain Science and Brain Computer Interface Technology School of Electrical and Information Engineering Zhengzhou University China School of Civil and Hydraulic Engineering Huazhong University of Science and Technology Wuhan430074 China

Objective: An electroencephalogram (EEG)-based brain-computer interface (BCI) enables direct communication between the human brain and a computer. Due to individual differences and non-stationarity of EEG signals, such BCIs usually require a subject-specific calibration session before each use, which is time-consuming and user-unfriendly. Transfer learning (TL) has been proposed to shorten or eliminate this calibration, but existing TL approaches mainly consider offline settings, where all unlabeled EEG trials from the new user are available. Methods: This paper proposes Test-Time Information Maximization Ensemble (T-TIME) to accommodate the most challenging online TL scenario, where unlabeled EEG data from the new user arrive in a stream, and immediate classification is performed. T-TIME initializes multiple classifiers from the aligned source data. When an unlabeled test EEG trial arrives, T-TIME first predicts its labels using ensemble learning, and then updates each classifier by conditional entropy minimization and adaptive marginal distribution regularization. Our code is publicized. Results: Extensive experiments on three public motor imagery based BCI datasets demonstrated that T-TIME outperformed about 20 classical and state-of-the-art TL approaches. Significance: To our knowledge, this is the first work on test time adaptation for calibration-free EEG-based BCIs, making plug-and-play BCIs possible. © 2024, CC BY.

关键词： Calibration