Zhixiang Wei

Zhixiang Wei

Auto-generated profile, based on publicly available information

Zhixiang Wei

School of Engineering Science, University of Science and Technology of China, Hefei, Anhui, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
29 December 2025 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
ZGZiyang Gong
Ziyang Gong
ZWZhixiang Wei
Zhixiang Wei
DWDi Wang
Di Wang
XHXiaoxing Hu
Xiaoxing Hu
XMXianzheng Ma
Xianzheng Ma
Liangpei ZhangLiangpei Zhang
Liangpei Zhang
Due to the substantial domain gaps in Remote Sensing (RS) images that are characterized by variabilities such as location, wavelength, and sensor type, Remote Sensing Domain Generalization (RSDG) has emerged as a critical and valuable research frontier, focusing on developing models that generalize effectively across diverse scenarios. However, research in this area remains underexplored: (1) Current cross-domain methods primarily focus on Domain Adaptation (DA), which adapts models to predefined domains rather than to unseen ones; (2) Few studies target the RSDG issue, especially for semantic segmentation tasks. Existing related models are developed for specific unknown domains, struggling with issues of underfitting on other unseen scenarios; (3) Existing RS foundation models tend to prioritize in-domain performance over cross-domain generalization. To this end, we introduce the first vision foundation model for RSDG semantic segmentation, CrossEarth. CrossEarth demonstrates strong cross-domain generalization through a specially designed data-level Earth-Style Injection pipeline and a model-level Multi-Task Training pipeline. In addition, for the semantic segmentation task, we have curated an RSDG benchmark comprising 32 semantic segmentation scenarios across various regions, spectral bands, platforms, and climates, providing comprehensive evaluations of the generalizability of future RSDG models. Extensive experiments on this collection demonstrate the superiority of CrossEarth over existing state-of-the-art methods.

1 August 2026 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
ZWZhixiang Wei
Zhixiang Wei
XMXiaoxiao Ma
Xiaoxiao Ma
RYRuishen Yan
Ruishen Yan
TTTao Tu
Tao Tu
HCHuaian Chen
Huaian Chen
ECEnhong Chen
Enhong Chen
Vision Foundation Models(VFMs) have achieved remarkable success in various computer vision tasks. However, their application to semantic segmentation is hindered by two significant challenges: (1) the disparity in data scale, as segmentation datasets are typically much smaller than those used for VFM pre-training, and (2) domain distribution shifts, where real-world segmentation scenarios are diverse and often underrepresented during pre-training. To overcome these limitations, we present Rein++, an efficient VFM-based segmentation framework that demonstrates superior generalization from limited data and enables effective adaptation to diverse unlabeled scenarios. Specifically, Rein++ comprises a domain generalization solution Rein G and a domain adaptation solution Rein-A. Rein-G introduces a set of trainable, instance-aware tokens that effectively refine the VFM's features for the segmentation task. This parameter efficient approach fine-tunes less than 1% of the backbone's parameters, enabling robust generalization. Building on the Rein-G, Rein-A performs unsupervised domain adaptation at both the instance and logit levels to mitigate domain shifts. In addition, it incorporates a semantic transfer module that leverages the class-agnostic capabilities of the segment anything model to enhance boundary details in the target domain. The integrated Rein++ pipeline first learns a generalizable model on a source domain (e.g., daytime scenes) and subsequently adapts it to diverse target domains (e.g., nighttime scenes) without any target labels. Comprehensive experiments demonstrate that Rein++ significantly outperforms state-of-the-art methods with efficient training, underscoring its roles an efficient, generalizable, and adaptive segmentation solution for VFMs, even for large models with billions of parameters. The code is available at https://github.com/wloves/Rein.

8 July 2025 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
Verified
Journal Article
Closed access
HCHuaian Chen
Huaian Chen
TLTianle Liu
Tianle Liu
BWBen Wang
Ben Wang
ZWZhixiang Wei
Zhixiang Wei
Yi JinYi Jin
Yi Jin
ECEnhong Chen
Enhong Chen
Imaging systems working under poor lighting conditions often suffer from poor visibility, compromising the reliability of vision-based systems. To address this problem, recent studies have developed data-driven low-light image enhancement (LIE) techniques to improve visibility. However, these LIE methods typically require large amounts of training samples, and the learned representation may not be valid for real-world scenes due to data discrepancies. In this work, we propose DP-LIE, an unsupervised LIE method driven by both data and priors. Unlike the existing methods that learn unexplainable high-dimensional features for end-to-end mapping, DP-LIE focuses on learning prior-guided parametric maps with definite meanings, enabling the low-light images to be brightened from an interpretable prior-based perspective. To this end, we design a simple yet effective prior-guided network-assisted LIE formulation, which elaborately bonds the data-driven representations with the traditional priors. The embedded priors narrow the solution space of the LIE model, allowing it to be efficiently trained with fewer samples. Notably, even trained with just a single low-light input image, the proposed method (denoted as DP-LIE-S) achieves comparable performances with existing unsupervised LIE methods. Moreover, experiments demonstrate that the proposed DP-LIE method exhibits excellent generalization performance across diverse imaging devices and promises better detection results for vision-based detection systems in nighttime scenes.

14 October 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
WGWenchao Guan
Wenchao Guan
BWBen Wang
Ben Wang
XMXiaoxiao Ma
Xiaoxiao Ma
JDJunkang Dai
Junkang Dai
CKChengfeng Kong
Chengfeng Kong
HCHuaian Chen
Huaian Chen
Domain bridging aims to eliminate discrepancies across domains by establishing an intermediary domain, enabling gradual domain adaptation. However, in the realm of video segmentation, existing methods often yield inaccurate motion within this intermediary domain, especially when tackling complex temporal sequences involving both object and camera movements. To overcome this challenge, we introduce a Motion-guided Domain Bridge (MDBridge), which constructs an aligned video intermediary domain and strengthens motion awareness. Central to MDBridge is the Motion-Aligned Domain Bridge , which unifies the relative camera movements into a consistent ego-motion by leveraging depth and motion priors. Complementing this, the Motion Perception Module , which utilizes monocular and binocular vision to establish a spatial perception structure, is developed to capture and utilize aligned motion information. Notably, both motion and depth information are derived using a self-supervised depth estimation strategy, without the need for external depth ground truth, allowing us to directly integrate into existing tasks without extra annotation costs. These progressive steps collectively extract and leverage temporal information from complex temporal scenarios, ensuring smooth domain transition during video segmentation training. Extensive experiments validate the superiority of our approach over previous state-of-the-art methods.

9 November 2026
Verified
Conference Paper
Open access
SLShifeng Liu
Shifeng Liu
ZZZhengye Zhang
Zhengye Zhang
SZSirui Zhao
Sirui Zhao
XMXinglong Mao
Xinglong Mao
ZKZhehan Kan
Zhehan Kan
ECEnhong Chen
Enhong Chen

14 January 2026 in Research Square
Verified
Preprint
Open access
PDF file
HCHuaian Chen
Huaian Chen
SHShiyao Hong
Shiyao Hong
YGYuxuan Gu
Yuxuan Gu
JDJunkang Dai
Junkang Dai
ZWZhixiang Wei
Zhixiang Wei
ECEnhong Chen
Enhong Chen
Fluorescence microscopy, empowered by artificial intelligence (AI), has shown great promise in advancing life science. However, the high cost and complexity of sample preparation limit the availability of training data, constraining the performance of AI-based models. Here, we present FluoGen, a diffusion-based generative foundation model designed to improve AI-based fluorescence image processing under data-constrained conditions. By pretraining on 3.5 million fluorescence images with a reformulated learning objective, FluoGen learns rich biological representations and mitigates inherent biases in conventional diffusion models. We demonstrate that FluoGen can serve as a backbone for image enhancement, enabling models to recover cellular and subcellular structures with substantially limited samples. Furthermore, FluoGen can reduce the training data required by existing AI-based analysis models to approximately 2% of the original amount while boosting the performance of state-of-the-art methods without architectural modifications. We anticipate that FluoGen serves as a foundation tool for advancing AI applications in fluorescence imaging.