Author name cluster

Mei Li

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

11 papers

2 author rows

EAAI Journal 2026 Journal Article

Leveraging multi-strategy labels for long-tailed classification

Wu Zeng
Mei Li

Most datasets in the real world predominantly exhibit long-tailed distribution characteristics. This inter-class imbalance phenomenon in datasets causes trained models to focus excessively on majority-class samples. Meanwhile, the models exhibit significantly poorer detection performance for minority-class samples. Although existing data augmentation methods can generate new samples that incorporate minority-class pixel regions. This approach improves the model’s recognition performance for minority-class samples. However, these methods typically assign labels to new samples solely based on the proportional area of pixel coverage between samples. This practice of ignoring semantic correlations among new samples often results in generated labels that substantially deviate from actual sample content, thereby providing erroneous supervisory signals that ultimately misguide the model training process. To address this issue, this paper proposes a multi-strategy mixed label with semantic awareness for recognition (MMLSAR) method. Specifically, we first construct a semantic predictor connected to the backbone network to more accurately assess the semantic composition of augmented images. Secondly, we propose an adaptive semantic adjustment module that enables dynamic allocation of label values. This allocation is achieved by integrating geometric features of cropped regions from augmented samples with semantic information from sample content. Finally, we design an augmented-sample label prediction auxiliary task. This task provides superior supervisory signals to the model by improving the prediction accuracy of augmented sample labels. We validated the effectiveness of our method on multiple datasets. In engineering applications, this technology can provide technical support to address class imbalance issues in real-world scenarios.

Details DOI

YNIMG Journal 2026 Journal Article

Social status modulates cooperative feedback processing: Electrophysiological evidence from an event-related potential study

Mei Li
Wenbin Pan
Xukai Zhang
Jialu Li
Jin Li
Qingfeng Peng
Hong Li

Collaboration is essential, as both one's own and others' performance impact outcomes. However, little is known about how social status affects performance and reward processing during cooperation. This study used event-related potentials (ERPs) to investigate neural responses during cooperative tasks with high- and low-status partners, where participants were assigned middle status after a math competition. ERP signals were recorded as they observed self-performance, partner performance, and cooperative outcomes. Results revealed asymmetric processing: participants referenced their performance when evaluating others', while self-performance processing was less influenced by others'. In the self-other performance order, participants showed more negative feedback-related negativity (FRN) following others' incorrect performance after their own correct performance, and larger P2 to others' correct performance after their own incorrect performance. In the other-self order, smaller P2 and more negative FRN were found for self-incorrect performance regardless of partners' performance, with only the P2 effect larger when partners were incorrect. Moreover, high-status partners elicited more negative FRN than low-status partners when others performed correctly; this difference disappeared when others performed incorrectly. For order-independent cooperative outcomes, with high-status partners, the FRN was more negative and the P3a was smaller for self-loss than self-gain when partners gained, reversing when partners lost. With low-status partners, only the P3a difference persisted when partners gained, while both components mirrored the high-status pattern when partners lost. These findings suggest that social status shapes sequential cooperative feedback processing, particularly during the later integrative stage in which cooperative outcomes are evaluated under prior knowledge of self and partner performance.

Details DOI

EAAI Journal 2026 Journal Article

Surging early warning in gas-liquid two-phase centrifugal pump based on autoencoder with contrastive learning

Quanrui Zhao
Denghui He
Suting Li
Xuwei Xiang
Mei Li
Bingwen Gao

Centrifugal pumps operating under gas-liquid two-phase flow conditions are highly susceptible to performance collapse due to surging, characterized by catastrophic drops in head, efficiency, and flow rate. Traditional surging prediction methods—relying on experimental correlations, computational fluid dynamics, or threshold-based monitoring—exhibit significant limitations in accuracy, real-time capability, and generalizability. To address this gap, this study proposes an intelligent early warning framework integrating a One-dimensional Convolutional Neural Network-Bidirectional Gated Recurrent Unit (1DCNN-BiGRU) autoencoder with contrastive learning. Pressure pulsation data were acquired from a transparent centrifugal pump across varied inlet gas volume fractions, rotational speeds, and seven monitoring positions. The baseline autoencoder, trained exclusively on non-surging data and optimized via Bayesian hyperparameter tuning, achieved effective reconstruction of normal operation patterns (max error <3 %). To enhance sensitivity near instability thresholds, a contrastive learning strategy was introduced: encoder weights were frozen, limited surging samples were incorporated, and a novel contrastive loss function with a dynamic threshold ε was designed to cluster normal samples while distancing surging representations. Results demonstrate that the contrastive model significantly outperformed the baseline, reducing the missed detection rate (MDR) by 17. 3 % on average across all sensor locations. Optimal warning performance occurred at the volute tongue (Position P3), achieving an F1-score of 0. 937, false alarm rate (FAR) of 5. 8 %, MDR of 6. 8 %, and response time of 0. 347 s. This work establishes a new paradigm for real-time surging prognostics, with direct implications for reliability in energy, petrochemical, and water infrastructure systems.

Details DOI

ICML Conference 2025 Conference Paper

BECAME: Bayesian Continual Learning with Adaptive Model Merging

Mei Li
Yuxiang Lu
Qinyan Dai
Suizhi Huang
Yue Ding 0001
Hongtao Lu 0001

Continual Learning (CL) strives to learn incrementally across tasks while mitigating catastrophic forgetting. A key challenge in CL is balancing stability (retaining prior knowledge) and plasticity (learning new tasks). While representative gradient projection methods ensure stability, they often limit plasticity. Model merging techniques offer promising solutions, but prior methods typically rely on empirical assumptions and carefully selected hyperparameters. In this paper, we explore the potential of model merging to enhance the stability-plasticity trade-off, providing theoretical insights that underscore its benefits. Specifically, we reformulate the merging mechanism using Bayesian continual learning principles and derive a closed-form solution for the optimal merging coefficient that adapts to the diverse characteristics of tasks. To validate our approach, we introduce a two-stage framework named BECAME, which synergizes the expertise of gradient projection and adaptive merging. Extensive experiments show that our approach outperforms state-of-the-art CL methods and existing merging strategies https: //github. com/limei0818/BECAME.

Details

ICML Conference 2024 Conference Paper

Sample Average Approximation for Conditional Stochastic Optimization with Dependent Data

Yafei Wang
Bo Pan
Mei Li
Jianya Lu
Lingchen Kong
Bei Jiang
Linglong Kong

Conditional Stochastic Optimization (CSO) is a powerful modelling paradigm for optimization under uncertainty. The existing literature on CSO is mainly based on the independence assumption of data, which shows that the solution of CSO is asymptotically consistent and enjoys a finite sample guarantee. The independence assumption, however, does not typically hold in many important applications with dependence patterns, such as time series analysis, operational control, and reinforcement learning. In this paper, we aim to fill this gap and consider a Sample Average Approximation (SAA) for CSO with dependent data. Leveraging covariance inequalities and independent block sampling technique, we provide theoretical guarantees of SAA for CSO with dependent data. In particular, we show that SAA for CSO retains asymptotic consistency and a finite sample guarantee under mild conditions. In addition, we establish the sample complexity $O(d / \varepsilon^4)$ of SAA for CSO, which is shown to be of the same order as independent cases. Through experiments on several applications, we verify the theoretical results and demonstrate that dependence does not degrade the performance of the SAA approach in real data applications.

Details

EAAI Journal 2023 Journal Article

An inverse halftoning method for various types of halftone images based on multi-scale generative adversarial network

Erhu Zhang
Mei Li
Qing Zhang
Lele Wu
Linhao Shao

The inverse halftoning method based on deep neural network has been widely studied in last few years. However, the existing methods only apply single-size convolution kernel to extract image features and train the proposed model with only one-scale loss function, which leads to blurring details and residual halftone noise patterns in the restored continuous-tone image. To alleviate these problems, this paper proposes an inverse halftone method for various types of halftone images based on multi-scale generative adversarial network. In the generative model, an elaborated multi-scale feature extraction model is firstly designed to extract halftone image features in parallel, so as to obtain more comprehensive image information from the input halftone image and promote the effective fusion of the extracted features. Moreover, a detail enhancement subnetwork is employed to fine-tune the image details. In adversarial model, a multi-scale discriminator is used to further enhance the image details by learning the distribution characteristics of image information. The experiments on different types of halftone images show that the proposed method significantly outperforms the existing state-of-art methods.

Details DOI

AIIM Journal 2023 Journal Article

Differential diagnosis of secondary hypertension based on deep learning

Lin Wu
Liying Huang
Mei Li
Zhaojun Xiong
Dinghui Liu
Yong Liu
Suzhen Liang
Hua Liang

Secondary hypertension is associated with higher risks of target organ damage and cardiovascular and cerebrovascular disease events. Early aetiology identification can eliminate aetiologies and control blood pressure. However, inexperienced doctors often fail to diagnose secondary hypertension, and comprehensively screening for all causes of high blood pressure increases health care costs. To date, deep learning has rarely been involved in the differential diagnosis of secondary hypertension. Relevant machine learning methods cannot combine textual information such as chief complaints with numerical information such as the laboratory examination results in electronic health records (EHRs), and the use of all features increases health care costs. To reduce redundant examinations and accurately identify secondary hypertension, we propose a two-stage framework that follows clinical procedures. The framework carries out an initial diagnosis process in the first stage, on which basis patients are recommended for disease-related examinations, followed by differential diagnoses of different diseases based on the different characteristics observed in the second stage. We convert the numerical examination results into descriptive sentences, thus blending textual and numerical characteristics. Medical guidelines are introduced through label embedding and attention mechanisms to obtain interactive features. Our model was trained and evaluated using a cross-sectional dataset containing 11, 961 patients with hypertension from January 2013 to December 2019. The F1 scores of our model were 0. 912, 0. 921, 0. 869 and 0. 894 for primary aldosteronism, thyroid disease, nephritis and nephrotic syndrome and chronic kidney disease, respectively, which are four kinds of secondary hypertension with high incidence rates. The experimental results show that our model can powerfully use the textual and numerical data contained in EHRs to provide effective decision support for the differential diagnosis of secondary hypertension.

Details DOI

YNIMG Journal 2022 Journal Article

A novel technology for in vivo detection of cell type-specific neural connection with AQP1-encoding rAAV2-retro vector and metal-free MRI

Ning Zheng
Mei Li
Yang Wu
Challika Kaewborisuth
Zhen Li
Zhu Gui
Jinfeng Wu
Aoling Cai

A mammalian brain contains numerous neurons with distinct cell types for complex neural circuits. Virus-based circuit tracing tools are powerful in tracking the interaction among the different brain regions. However, detecting brain-wide neural networks in vivo remains challenging since most viral tracing systems rely on postmortem optical imaging. We developed a novel approach that enables in vivo detection of brain-wide neural connections based on metal-free magnetic resonance imaging (MRI). The recombinant adeno-associated virus (rAAV) with retrograde ability, the rAAV2-retro, encoding the human water channel aquaporin 1 (AQP1) MRI reporter gene was generated to label neural connections. The mouse was micro-injected with the virus at the Caudate Putamen (CPU) region and subjected to detection with Diffusion-weighted MRI (DWI). The prominent structure of the CPU-connected network was clearly defined. In combination with a Cre-loxP system, rAAV2-retro expressing Cre-dependent AQP1 provides a CPU-connected network of specific type neurons. Here, we established a sensitive, metal-free MRI-based strategy for in vivo detection of cell type-specific neural connections in the whole brain, which could visualize the dynamic changes of neural networks in rodents and potentially in non-human primates.

Details DOI

AAAI Conference 2022 Conference Paper

Noninvasive Lung Cancer Early Detection via Deep Methylation Representation Learning

Xiangrui Cai
Jinsheng Tao
Shichao Wang
Zhiyu Wang
Jiaxian Wang
Mei Li
Hong Wang
Xixiang Tu

Early detection of lung cancer is crucial for five-year survival of patients. Compared with the pathological analysis and CT scans, the circulating tumor DNA (ctDNA) methylation based approach is noninvasive and cost-effective, and thus is one of the most promising methods for early detection of lung cancer. Existing studies on ctDNA methylation data measure the methylation level of each region with a predefined metric, ignoring the positions of methylated CpG sites and methylation patterns, thus are not able to capture the early cancer signals. In this paper, we propose a blood-based lung cancer detection method, and present the first ever study to represent methylation regions by continuous vectors. Specifically, we propose DeepMeth to regard each region as a one-channel image and develop an auto-encoder model to learn its representation. For each ctDNA methylation sample, DeepMeth achieves its representation via concatenating the region vectors. We evaluate DeepMeth on a multicenter clinical dataset collected from 14 hospitals. The experiments show that DeepMeth achieves about 5%-8% improvements compared with the baselines in terms of Area Under the Curve (AUC). Moreover, the experiments also demonstrate that DeepMeth can be combined with traditional scalar metrics to enhance the diagnostic power of ctDNA methylation classifiers. DeepMeth has been clinically deployed and applied to 450 patients from 94 hospitals nationally since April 2020.

PDF Details

ICRA Conference 2021 Conference Paper

Direct Sparse Stereo Visual-Inertial Global Odometry

Ziqiang Wang
Mei Li
Dingkun Zhou
Ziqiang Zheng

Robust and accurate localization plays a key role in autonomous driving and robot applications. To utilize the complementary properties of different sensors, we present a novel tightly-coupled approach to combine the local (stereo cameras, IMU) and global sensors (magnetometer, GNSS). We jointly optimize all the model parameters through one active window. The visual part integrates constraints from static stereo into the photometric bundle adjustment pipeline of dynamic multiview stereo. Accumulating IMU information between keyframes, magnetometer and GNSS measurements are all inserted into the active window as additional constrains among all the keyframes. Through these, our method can realize globally drift-free and locally accurate state estimation. We evaluate the effectiveness of our system on public datasets under with real-world experiments.

Details

JBHI Journal 2021 Journal Article

Predicting Recurrence for Patients With Ischemic Cerebrovascular Events Based on Process Discovery and Transfer Learning

Haifeng Xu
Jianfei Pang
Weiliang Zhang
Xuemeng Li
Mei Li
Dongsheng Zhao

The recurrence of Ischemic cerebrovascular events (ICE) often results in a high rate of mortality and disability. However, due to the lack of labeled follow-up data in hospitals, prediction methods using traditional machine learning are usually not available or reliable. Therefore, we propose a new framework for predicting the long-term recurrence risk in patients with ICE after discharge from hospitals based on process mining and transfer learning, to point out high-risk patients for intervention. First, process models are discovered from clinical guidelines for analyzing the similarity of ICE population data collected by different medical institutions, and the control flow found are taken as added characteristics of patients. Then we use the in-hospital data (target domain) and the national stroke screening data (source domain), to develop risk prediction models applying instance filter and weight-based transfer learning method. To verify our method, 205 cases from a tertiary hospital and 2954 cases from the screening cohort (2015-2017) are tested. Experimental results show that our framework can improve the performance of three instance-based transfer algorithms. This study provides a comprehensive and efficient approach for applying transfer learning, to alleviate the limitation of insufficient labeled follow-up data in hospitals.

Details DOI