Chen Li

Dr. Chen Li

orcid

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

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

Representative Publications

View All Publications
22 June 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
KWKaiqi Wang
Kaiqi Wang
LJLiu Jianhua
Liu Jianhua
CLChen Li
Chen Li
BYBing Yu
Bing Yu
Accurate and automated extraction of solid waste sites from remote-sensing imagery constitutes a pivotal demand for contemporary environmental regulation and risk mitigation. However, in high-resolution remote-sensing imagery, solid waste sites are typically represented as a single semantic image object (SIO), which is composed of multiple physical image parcels (PIPs) exhibiting significant variations in scale, morphology, and spectral properties. This intrinsic heterogeneity substantially increases the complexity and uncertainty of multi-class site identification. To address this challenge, this paper proposes WasteMOE Net (WMN), which is developed based on the core concept of modeling the SIO–PIP relationship. WMN adopts a heterogeneous expert selection mechanism combined with a nested mixture-of-experts architecture. It thus enables adaptive perception of complex PIPs across diverse scenarios and their integrated discrimination at the SIO level. In addition, by incorporating the explicit nonlinear representation capability of the KAN network, WMN effectively improves multi-class recognition accuracy while maintaining computational efficiency. Furthermore, this study constructs a high-resolution solid waste site dataset in accordance with the SIO–PIP-aware annotation principle, encompassing five representative categories: tailings ponds (TP), construction spoil sites (CSS), landfill sites (LS), garbage dump sites (GDS), and excavation sites (ES). Experimental results show that WMN achieves mAP50 values of 74.2% (GDS), 63.5% (CSS), 80.9% (ES), 85.4% (TP), and 83.1% (LS) in detection tasks, and 75.4%, 64.1%, 83.0%, 86.7%, and 84.1% for the corresponding categories in segmentation tasks. It achieves competitive performance compared with state-of-the-art methods in both tasks. Further, in a real-world application over Loudi City, China, WMN completed the processing of a 490.67 km2 area within 1.34 h. The recognition accuracies for GDS and ES reached 54.8% and 65.3%, respectively. Finally, the proposed method has been successfully integrated into a GIS-based solid waste pollution risk prevention system, which markedly boosts the overall efficiency of environmental monitoring and on-site inspections.

4 November 2026 in Measurement
Verified
Journal Article
Closed access
CLChen Li
Chen Li
JZJiannan Zhang
Jiannan Zhang
XWXinyu Wang
Xinyu Wang
KWKaiqi Wang
Kaiqi Wang
LJLiu Jianhua
Liu Jianhua

30 August 2026 in Sensors
Verified
Journal Article
Open access
PDF file
LJLiu Jianhua
Liu Jianhua
CLChen Li
Chen Li
XWXinyu Wang
Xinyu Wang
BWBozhen Wang
Bozhen Wang
Building facades are continually exposed to weathering, material aging, hygrothermal cycling, and service-related disturbances. Cracks, delamination, spalling, and seepage can compromise durability and safety, while associated thermal anomalies provide indirect evidence of deterioration. Deep learning, UAV inspection, infrared thermography, LiDAR, building information modeling (BIM), and digital twins have shifted facade inspection toward automated sensing and spatially referenced assessment. Yet most studies remain focused on isolated measures of model accuracy and give limited attention to the full pathway from detection and quantification to component localization, condition rating, and repair. This review organizes the evidence along five dimensions: annotation granularity (A0–A2/A2*), visual task hierarchy (T1–T5), fusion level (F0–F3), spatial mapping level (S0–S3), and engineering maturity (M1–M3). The synthesis indicates that acquisition conditions, material heterogeneity, negative-sample composition, and annotation rules constrain model generalization. Single-modality methods cover classification, detection, segmentation, and partial geometric quantification, but facade-specific external validation remains limited. Multimodal gains depend on registration quality, thermophysical conditions, defect mechanisms, and sensor reliability. Mapping two-dimensional outputs to point clouds, BIM, and digital twins is technically feasible, but error propagation, component matching, and condition-rating protocols remain inconsistent. Interpretability tools and vision foundation models may support verification, annotation, and reporting but cannot replace dedicated systems with validated error bounds. Future work should prioritize cross-dataset benchmarks, facade-specific lightweight models, reproducible multimodal evaluation, standardized 2D-to-3D accuracy protocols, uncertainty-aware manual review, and engineering condition-rating frameworks.