Yushi Hao

Yushi Hao

Auto-generated profile, based on publicly available information

Yushi Hao

School of Electrical and Information Engineering, Tianjin University, Tianjin, China

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
17 January 2025 in IEEE Sensors Journal
Verified
Journal Article
Closed access
JSJingyu Su
Jingyu Su
Shan AnShan An
Shan An
GWGuoxin Wang
Guoxin Wang
XSXinlin Sun
Xinlin Sun
YHYushi Hao
Yushi Hao
ZGZhongke Gao
Zhongke Gao
Due to the variability and non-stationarity of electroencephalography (EEG) signal across different recording scenarios and subjects, it is crucial to have methods with strong generalization capabilities that can effectively capture both temporal and spatial features while maintaining high accuracy. In this paper, we introduce TMCNet, a novel end-to-end deep learning model that integrates multi-scale 3D convolutional neural networks with Transformer for enhanced feature extraction. First, the temporal convolution is applied to EEG signals to extract detailed temporal features. Then we utilize multi-scale 3D convolutional branches to perform spatial convolution, capturing spatial information across diverse receptive fields. In the second stage, we utilize the multi-head attention mechanism in Transformer to extract more refined global features, with fully connected layers used for classification. We evaluate the proposed method on three publicly available datasets. In cases where the datasets contain information for two sessions, we conduct evaluations for both within-session and cross-session scenarios. The experimental results demonstrate that the proposed TMCNet exhibits advanced performance, showcasing strong decoding ability and robustness.

23 May 2025 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
XJXuheng Jiang
Xuheng Jiang
XSXinlin Sun
Xinlin Sun
YHYushi Hao
Yushi Hao
QCQing Cai
Qing Cai
HHHongjun Hou
Hongjun Hou
ZGZhongke Gao
Zhongke Gao
Electroencephalography (EEG) signals play a vital role in brain science research and clinical diagnostics. However, these signals are often contaminated by physiological artifacts such as electromyography (EMG), electrooculography (EOG), electrocardiography (ECG), and nonphysiological noise like white Gaussian noise (WGN), which substantially reduce the accuracy of analysis and decoding. In recent years, deep learning (DL) techniques have shown significant promise in EEG denoising tasks. However, existing methods struggle to capture the complex, nonlinear relationships inherent in EEG signals, particularly in multiscale feature extraction and time-frequency representation. To address these limitations, we propose WNOTNet, a novel model that integrates the Wavelet Neural Operator (WNO) and Transformer architecture. The WNO enables multiscale analysis and captures intricate time-frequency features, while the Transformer effectively models long-range dependencies and global context. This integration significantly improves both denoising performance and generalization capabilities. Extensive experiments conducted on the publicly available EEGdenoiseNet dataset show that WNOTNet outperforms existing methods across multiple evaluation metrics, including time-relative root mean square error (T-RRMSE), spectral-relative root mean square error (S-RRMSE), correlation coefficient (CC), signal-to-noise ratio (SNR), and others. In real-world applications, WNOT-Net was evaluated on practical EEG classification tasks, including sleep stage classification (MIT-BIH Polysomnographic dataset), driver fatigue detection, and emotion recognition. Notably, after denoising with WNOTNet, classification accuracy improved by 2.53% for sleep staging, 7.88% for driver fatigue detection, and 11.82% for emotion recognition. These experimental results demonstrate that WNOTNet can effectively enhance EEG signal quality and improve downstream classification accuracy. These findings indicate that WNOTNet provides robust and reliable denoising performance with strong generalization potential, making it suitable for practical EEG-based applications.

19 March 2025 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
XSXinlin Sun
Xinlin Sun
YHYushi Hao
Yushi Hao
Wei GuoWei Guo
Wei Guo
BJBochang Jiang
Bochang Jiang
HLHaoyu Li
Haoyu Li
ZGZhongke Gao
Zhongke Gao
Prolonged fatigue not only affects people’s learning and work efficiency, but also leads to a series of symptoms such as insomnia and forgetfulness. Timely and accurately identifying fatigue state is crucial in various industries. However, existing fatigue detection methods either rely on manually extracted features which is unable to fully utilize the deep level information of signals, or are complexly designed and hard to be implemented. In this paper, we propose spectral group-guided lightweight CNN (SGL-Net), which is an ultra-lightweight CNN model for fatigue detection. The design concept of SGL-Net is closely related to the mechanism of information processing in human brain. Firstly, the spectral space embedding decomposes the EEG signal into various rhythms, we also enrich the decomposition tree using wavelet convolution, where the complex rhythm information is decoupled. Secondly, we propose a novel spatial-temporal modality encoder, which captures the relationship among electrodes and evaluates the power spectrum of different rhythms. Finally, a kernel-restricted multi layer perceptron (MLP) is adopted for fatigue detection, ensuring the parameter sparsity simultaneously. We also design a well-suited hardware system for EEG acquisition on forehead. Experimental results have demonstrated the robustness, effectiveness, and practicability of the SGL-Net in real-world applications.

9 May 2025 in IEEE Sensors Journal
Verified
Journal Article
Closed access
BZBang Zhou
Bang Zhou
HCHanqing Chen
Hanqing Chen
MLMengyu Li
Mengyu Li
YHYushi Hao
Yushi Hao
ZLZezhong Li
Zezhong Li
ZGZhongke Gao
Zhongke Gao
Oil–water two-phase flow is a common phenomenon in industrial processes. Precisely measuring its parameters is crucial for monitoring production conditions and optimizing production strategies. Due to the complexity of oil–water two-phase flow, the precise simultaneous measurements of water cut and total flow rate are challenging tasks. To address this issue, we conduct vertical upward oil–water two-phase flow experiments using a dual-helix microwave sensor measurement system to collect fluid signals and design a dual-parameter measurement model (DPMM) based on multi-task learning for the simultaneous prediction of water cut and total flow rate. Firstly, DPMM utilizes the shared feature extractor (SFE) to extract shared features, facilitating interaction across tasks. Then, DPMM employs the multi-head attention network (MHAN) and the expanding kernel network (EKN) to extract specific features for water cut and total flow rate, respectively. Furthermore, a joint loss function is used to achieve collaborative optimization. Experimental results show that DPMM demonstrates outstanding performance in two-parameter measurements and surpasses other competing models. We verified the efficiency of the proposed methods by parameter analysis and ablation experiments.

6 May 2026 in Chaos, Solitons & Fractals
Verified
Journal Article
Closed access
YHYushi Hao
Yushi Hao
XJXuheng Jiang
Xuheng Jiang
QCQing Cai
Qing Cai
XGXiang Gao
Xiang Gao
JAJinapeng An
Jinapeng An
ZGZhongke Gao
Zhongke Gao

16 September 2026 in Journal of Neural Engineering
Verified
Journal Article
Closed access
JHJinchen Hu
Jinchen Hu
ZGZhongke Gao
Zhongke Gao
YHYushi Hao
Yushi Hao
SXShangbin Xie
Shangbin Xie
XLXiaoyang Li
Xiaoyang Li
JAJinapeng An
Jinapeng An
Objective. Electroencephalography (EEG) is widely used in brain-computer interfaces, but its microvolt-level signals are easily contaminated by electromyography (EMG), electrooculography (EOG), and mixed physiological artifacts. This study develops an EEG denoising framework that suppresses artifacts while preserving information used by downstream biomedical artificial intelligence tasks. Approach. We propose MTGNet, a task-oriented and spectrally guided EEG denoising framework. MTGNet combines low-rank adaptation (LoRA)-based task-aware consistency regularization, a spectrally aware Guidance Network, and parallel Mamba–Transformer backbone. A pretrained 11.97 M-parameter backbone learns to preserve intrinsic EEG characteristics from paired noisy-clean data, while 0.33 M LoRA parameters enable task-specific adaptation without requiring paired clean EEG references. Main results. On EEGDenoiseNet, MTGNet reduces spectral relative root-mean-square error by over 18.9%, 31.5%, and 14.0% for EMG, EOG, and hybrid artifacts, respectively ( p < 0.001). On a real-world fatigue EEG dataset, it improves classification accuracy by 6.20–6.69 percentage points compared with unprocessed inputs ( p < 0.05). Ablation, cross-classifier, and cross-dataset analyses validate the proposed components and support the transferability of MTGNet across the evaluated settings. Significance. The proposed framework provides a practical approach to task-aware EEG denoising, while future work should further validate its applicability across real-world EEG settings involving diverse tasks, artifact types, and acquisition conditions.