Ye Xu

Dr. Ye Xu

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 June 2020 in PLOS ONE
Verified
Journal Article
Open access
PDF file
YXYe Xu
Ye Xu
XYXiaodong Yu
Xiaodong Yu
TWTian Wang
Tian Wang
ZXZezhong Xu
Zezhong Xu
In the problem where there is not enough data to use Deep Learning, Bag-of-Visual-Words (BoVW) is still a good alternative for image classification. In BoVW model, many pooling methods are proposed to incorporate the spatial information of local feature into the image representation vector, but none of the methods devote to making each visual word have its own pooling regions. The practice of designing the same pooling regions for all the words restrains the discriminability of image representation, since the spatial distributions of the local features indexed by different visual words are not same. In this paper, we propose to make each visual word have its own pooling regions, and raise a simple yet effective method for learning pooling region. Concretely, a kind of small window named observation window is used to obtain its responses to each word over the whole image region. The pooling regions of each word are organized by a kind of tree structure, in which each node indicates a pooling region. For each word, its pooling regions are learned by constructing a tree with its labelled coordinate data. The labelled coordinate data consist of the coordinates of responses and image class labels. The effectiveness of our method is validated by observing if there is an obvious classification accuracy improvement after applying our method. Our experimental results on four small datasets (i.e., Scene-15, Caltech-101, Caltech-256 and Corel-10) show that, the classification accuracy is improved by about 1% to 2.5%. We experimentally demonstrate that the practice of making each word have its own pooling regions is beneficial to image classification task, which is the significance of our work.

29 February 2024 in PLOS ONE
Verified
Journal Article
Open access
PDF file
YXYe Xu
Ye Xu
XZXin Zhang
Xin Zhang
CHChongpeng Huang
Chongpeng Huang
XQXiaorong Qiu
Xiaorong Qiu
This article investigates whether higher classification accuracy can always be achieved by utilizing a pre-trained deep learning model as the feature extractor in the Bag-of-Deep-Visual-Words (BoDVW) classification model, as opposed to directly using the new classification layer of the pre-trained model for classification. Considering the multiple factors related to the feature extractor -such as model architecture, fine-tuning strategy, number of training samples, feature extraction method, and feature encoding method—we investigate these factors through experiments and then provide detailed answers to the question. In our experiments, we use five feature encoding methods: hard-voting, soft-voting, locally constrained linear coding, super vector coding, and fisher vector (FV). We also employ two popular feature extraction methods: one (denoted as Ext-DFs(CP)) uses a convolutional or non-global pooling layer, and another (denoted as Ext-DFs(FC)) uses a fully-connected or global pooling layer. Three pre-trained models—VGGNet-16, ResNext-50(32×4d), and Swin-B—are utilized as feature extractors. Experimental results on six datasets (15-Scenes, TF-Flowers, MIT Indoor-67, COVID-19 CXR, NWPU-RESISC45, and Caltech-101) reveal that compared to using the pre-trained model with only the new classification layer re-trained for classification, employing it as the feature extractor in the BoDVW model improves the accuracy in 35 out of 36 experiments when using FV. With Ext-DFs(CP), the accuracy increases by 0.13% to 8.43% (averaged at 3.11%), and with Ext-DFs(FC), it increases by 1.06% to 14.63% (averaged at 5.66%). Furthermore, when all layers of the pre-trained model are fine-tuned and used as the feature extractor, the results vary depending on the methods used. If FV and Ext-DFs(FC) are used, the accuracy increases by 0.21% to 5.65% (averaged at 1.58%) in 14 out of 18 experiments. Our results suggest that while using a pre-trained deep learning model as the feature extractor does not always improve classification accuracy, it holds great potential as an accuracy improvement technique.

23 December 2023 in Multimedia Tools and Applications
Verified
Journal Article
Closed access
PDF file
YXYe Xu
Ye Xu
LDLihua Duan
Lihua Duan
CHConggui Huang
Conggui Huang
CHChongpeng Huang
Chongpeng Huang

20 February 2026 in Journal of Vibroengineering
Verified
Journal Article
Open access
PDF file
XQXiaorong Qiu
Xiaorong Qiu
YXYe Xu
Ye Xu
In this paper, a simulation data-driven intelligent fault diagnosis algorithm based on attention mechanism and transfer learning is proposed to address insufficient fault data, low diagnostic accuracy, and inefficiency in rolling bearing monitoring under cross-condition and cross-location scenarios. To overcome the lack of real fault data, a dynamic vibration-response model is constructed through analysis of bearing and fault dynamics, generating high-fidelity fault signals across multiple operating conditions. Based on this, a diagnostic model is developed using a self-attention–assisted weighted autoencoder, where the proposed weighted autoencoder integrates a self-attention mechanism and a weight allocation mechanism and the former captures inter-feature dependencies while the latter adaptively reweighting feature contributions to enhance fault-discriminative representations. Therefore, the diagnostic model can assign corresponding weights to different importance features according to the constructed self-attention mechanism-assisted weighted self-encoding feature extraction model, effectively avoiding the problems of insufficient diagnostic accuracy and low diagnostic efficiency caused by feature redundancy and difficulty in distinguishing the importance of rolling bearing faults. Furthermore, the local maximum mean discrepancy (LMMD) method is applied to align both global and sub-domain distributions between simulated and measured data. By synthesizing cross-condition and cross-location simulated signals with real measurements, an LMMD-based intelligent transfer diagnosis model is built to enhance generalization and robustness against large distribution discrepancies. Finally, the stability and robustness of the proposed method are validated by analyzing the transfer learning performance and anti-noise disturbance ability across different operating conditions and locations.

1 November 2017
Verified
Conference Paper
Closed access
YXYe Xu
Ye Xu
XYXiaodong Yu
Xiaodong Yu
TWTian Wang
Tian Wang
FLF Lu
F Lu
Spatial pyramid is a very popular method for preserving the spatial information of local features, which partitions an image into multiple blocks at different resolution levels. Nevertheless, the strategy for partitioning an image is designed by hand and same for all the codewords in a codebook. To address this problem, we propose a novel partition method named discriminative spatial tree (ST), and implement it from two different viewpoints: class and image. Discriminative ST builds one spatial tree for each codeword, and each node in each tree corresponds to one image region. For better performance evaluation, we also introduce a simplified coding scheme. Experimental results on two challenging datasets show that our simplified coding scheme leads to comparable results to some sophisticated ones, and discriminative ST can achieve better classification performance than spatial pyramid.

9 September 2020 in IEEE Access
Verified
Journal Article
Open access
YXYe Xu
Ye Xu
LDLihua Duan
Lihua Duan
XYXiaodong Yu
Xiaodong Yu
TWTian Wang
Tian Wang
YSYingzhong Shi
Yingzhong Shi
Bag-of-Visual-Words (BoVW) is still a useful image classification model when there is not enough data to use Deep Learning. In BoVW model, the practice of reducing the reconstruction errors of local features can improve the classification accuracy owing to the decrease of information loss. Many reconstruction-based coding methods are proposed to learn a visual dictionary and encode local features via minimizing the reconstruction errors of local features with constraints. Besides this, the accuracy can also be improved by learning the category-specific dictionaries and then encoding features based on these dictionaries. By considering the two practices together, we propose a simple category-specific dictionary learning method tailored for reconstruction-based feature coding. Our method can be used as a universal one to improve the classification accuracies of many reconstruction-based coding methods, which is the highlight of our method. Concretely, a universal dictionary is learned by employing a reconstruction-based coding method and then refined for each category to obtain the category-specific dictionary of this category. When encoding a feature by a category-specific dictionary, the visual words for encoding it are decided in advance by the indices, which correspond to the non-zero elements of its coding vector obtained with the universal dictionary. The effectiveness of our method is validated by observing whether there is an accuracy improvement after applying our method. Our results on Scene-15, Caltech-101, and UIUC-Sports datasets show that the accuracies of four representative coding methods are improved by about 0.3% to 2.7%, which experimentally demonstrates the universality and effectiveness of our method.