Yi Xu

Dr. Yi Xu


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

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

Representative Publications

View All Publications
20 February 2024 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
RXRui Xu
Rui Xu
XDXue-Mei Dong
Xue-Mei Dong
WLWeijie Li
Weijie Li
Jiangtao PengJiangtao Peng
Jiangtao Peng
Weiwei SunWeiwei Sun
Weiwei Sun
YXYi Xu
Yi Xu
Currently, deep learning methods represented by convolutional neural networks (CNNs) or Transformers are of great interest in hyperspectral image (HSI) classification. And recent works show that hybrid models using CNN and Transformer modules are expected to achieve better performance than when they are used alone. However, these hybrid models applied to HSI classification consider the combination of 2D CNN and Transformer, which makes the models have high computational complexity. And the information of multiple spectral dimensions different from ordinary RGB images has not been fully excavated. Based on this, we propose DBCTNet, a double branch Convolution-Transformer network. Specifically, a MSpeFE module is used for multiscale spectral feature extraction at the early stage of the proposed network. Then a ConvTE block is designed to improve the original Transformer encoder, where a Conv spectral projection unit and a convolutional multihead self-attention (CMHSA) unit are proposed to extract spatial and global spectral features. A double branch module is further built based on 3D CNN and ConvTE. This module can fully integrate spatial and local-global spectral features, while also having low computational complexity. Experiment results on four public datasets, Pavia University, Houston, WHU-Hi-LongKou and HuangHeKou, show that DBCTNet achieves satisfactory performance with a small number of parameters and relatively excellent efficiency compared to nine other networks. The implement of DBCTNet will be available publicly at https://github.com/xurui-joei/DBCTNet.

30 October 2024
Verified
Book Chapter
Closed access
PDF file
QZQiran Zou
Qiran Zou
SYShangyuan Yuan
Shangyuan Yuan
SDShian Du
Shian Du
YWYu Wang
Yu Wang
CLChang Liu
Chang Liu
XJXiangyang Ji
Xiangyang Ji

22 May 2024 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
YYYang Yang
Yang Yang
NJNan Jiang
Nan Jiang
YXYi Xu
Yi Xu
DZDe-Chuan Zhan
De-Chuan Zhan
Open-set Semi-supervised Learning (OSSL) holds a realistic setting that unlabeled data may come from classes unseen in the labeled set, i.e., out-of-distribution (OOD) data, which could cause performance degradation in conventional SSL models. To handle this issue, except for the traditional in-distribution (ID) classifier, some existing OSSL approaches employ an extra OOD detection module to avoid the potential negative impact of the OOD data. Nevertheless, these approaches typically employ the entire set of open-set data during their training process, which may contain data unfriendly to the OSSL task that can negatively influence the model performance. This inspires us to develop a robust open-set data selection strategy for OSSL. Through a theoretical understanding from the perspective of learning theory, we propose Wise Open -set Semi-supervised Learning (WiseOpen), a generic OSSL framework that selectively leverages the open-set data for training the model. By applying a gradient-variance-based selection mechanism, WiseOpen exploits a friendly subset instead of the whole open-set dataset to enhance the model's capability of ID classification. Moreover, to reduce the computational expense, we also propose two practical variants of WiseOpen by adopting low-frequency update and loss-based selection respectively. Extensive experiments demonstrate the effectiveness of WiseOpen in comparison with the state-of-the-art.

18 April 2024 in Applied Intelligence
Verified
Journal Article
Closed access
PDF file
PDPeixuan Ding
Peixuan Ding
YXYi Xu
Yi Xu
PQPan Qin
Pan Qin
XSXi-Ming Sun
Xi-Ming Sun

30 April 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
PDPeixuan Ding
Peixuan Ding
YXYi Xu
Yi Xu
XSXiming Sun
Ximing Sun
Maintaining the operational stability of an aero-engine heavily relies on the health of the rotor bearings. Data-driven fault diagnosis approaches have achieved significant success due to their superior ability to identify faults. However, the application of conventional deep learning methods, which heavily rely on a huge amount of vibration signals, is unrealistic in most aero-engine-related cases due to the limited number of fault samples. To tackle this problem, a multitask Siamese network (MTSN) is proposed, which combines data augmentation (DA) and metric learning into a unified framework. First, in the multiclassification task, a limited number of samples are augmented using Mixup and learned by a convolutional neural network (CNN). Then, in the metric learning task, the similarity between sample pairs is learned by the shared encoder in the Siamese network. To ensure effective learning in both tasks, the loss functions of the two learning tasks are balanced by introducing uncertainty. Based on two public bearing fault datasets, we conducted a series of experiments. The experimental results show that the learning tasks mutually reinforce each other, and the MTSN exhibits good fault diagnosis performance with limited data.

6 March 2025 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
YYYang Yang
Yang Yang
HPHongpeng Pan
Hongpeng Pan
QJQing-Yuan Jiang
Qing-Yuan Jiang
YXYi Xu
Yi Xu
JTJinhui Tang
Jinhui Tang
Multi-modal learning aims to enhance performance by unifying models from various modalities but often faces the “modality imbalance” problem in real data, leading to a bias towards dominant modalities and neglecting others, thereby limiting its overall effectiveness. To address this challenge, the core idea is to balance the optimization of each modality to achieve a joint optimum. Existing approaches often employ a modal-level control mechanism for adjusting the update of each modal parameter. However, such a global-wise updating mechanism ignores the different importance of each parameter. Inspired by subnetwork optimization, we explore a uniform sampling-based optimization strategy and find it more effective than global-wise updating. According to the findings, we further propose a novel importance sampling-based, element-wise joint optimization method, called Adaptively Mask Subnetworks Considering Modal Significance (AMSS). Specifically, we incorporate mutual information rates to determine the modal significance and employ non-uniform adaptive sampling to select foreground subnetworks from each modality for parameter updates, thereby rebalancing multi-modal learning. Additionally, we demonstrate the reliability of the AMSS strategy through convergence analysis. Building upon theoretical insights, we further enhance the multi-modal mask subnetwork strategy using unbiased estimation, referred to as AMSS+. Extensive experiments reveal the superiority of our approach over comparison methods.