Wenzhi Cheng

Dr. Wenzhi Cheng


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
27 May 2025 in Forests
Verified
Journal Article
Open access
PDF file
Linhui WangLinhui Wang
Linhui Wang
Linhui WangLinhui Wang
Linhui Wang
XPXuxiang Peng
Xuxiang Peng
YTYonghong Tan
Yonghong Tan
ZZZhenqi Zhou
Zhenqi Zhou
XLXiaolin Liang
Xiaolin Liang
During the flowering period of mango trees, pests often hide in the inflorescences to suck sap, affecting fruit formation. By accurately detecting the number and location of mango inflorescences in the early stages, it can help target-specific spraying equipment to perform precise pesticide application. This study focuses on mango panicles and addresses challenges such as high crop planting density, poor image quality, and complex backgrounds. A series of improvements were made to the YOLOv8 model to enhance performance for this type of detection task. Firstly, a mango panicle dataset was constructed by selecting, augmenting, and correcting samples based on actual agricultural conditions. Second, the backbone network of YOLOv8 was replaced with FasterNet. Although this led to a slight decrease in accuracy, it significantly improved inference speed and reduced model parameters, demonstrating that FasterNet effectively reduced computational complexity while optimizing accuracy. Further, the GAM (Global Attention Module) attention mechanism was introduced as an attention module in the backbone network to enhance feature extraction capabilities. Experimental results indicated that the addition of GAM improved the average precision by 2.2 percentage points, outperforming other attention mechanisms such as SE, CA, and CBAM. Finally, the model’s bounding box localization ability was enhanced by replacing the loss function with WIoU, which also accelerated model convergence and improved the mAP@.5 metric by 1.1 percentage points. Our approach demonstrates a discrepancy of less than 10% compared to manual counted results.

8 January 2026 in Discover Applied Sciences
Verified
Journal Article
Open access
PDF file
LWLi Wan
Li Wan
WCWenzhi Cheng
Wenzhi Cheng
Retransmitting multimedia data files of varying lengths in wireless networks often suffers from inefficient coding, resulting in frequent retransmissions and reduced quality of service (QoS). Existing retransmission methods face two major limitations: traditional sequential splicing approaches struggle with overlapping packet losses in variable-length scenarios, and packet-length headers introduce substantial transmission overhead that degrades network performance. This study introduces BRAGNC, a wireless multimedia broadcast retransmission algorithm based on deep convolutional generative adversarial network (DCGAN) coding. BRAGNC addresses these limitations through three key mechanisms: a segment-by-segment approach with enhanced flag registers to minimize data processing time, adaptive packet splicing based on node feedback to improve retransmission efficiency, and flag bit delimiters to reduce header overhead. Experiments show that BRAGNC reduces retransmission frequency by 25%, improves coding efficiency by 15%, and decreases network overhead by 20% across diverse packet loss scenarios. These results provide a basis for advancing generative network coding in multimedia broadcasting.

25 September 2026 in Insects
Verified
Journal Article
Open access
PDF file
QLQinchun Li
Qinchun Li
FXFanping Xu
Fanping Xu
WCWenzhi Cheng
Wenzhi Cheng
CQChengmei Qi
Chengmei Qi
XLXiaowen Liu
Xiaowen Liu
Biological invasions are a critical driver of global environmental change, with plant–insect interactions central to invasion success and biological control. This review systematically examines three classical hypotheses—the Enemy Release Hypothesis (ERH), the Novel Weapons Hypothesis (NWH), and the Evolution of Increased Competitive Ability (EICA) hypothesis, within a plant–insect context. Our synthesis shows that ERH support is widespread but highly contingent on environmental context; NWH highlights the negative effects of invasive allelochemicals on native insects and soil biota; and the EICA hypothesis provides an evolutionary framework for understanding shifts in defence–growth resource allocation in invasive plants. We advocate for a multitrophic network perspective that captures ecosystem complexity and elucidates synergistic interactions among hypotheses across space and time. We identify inherent limitations in single-hypothesis biocontrol approaches and propose an integrative management framework grounded in network theory. Emerging tools—including molecular ecology and network modelling—are emphasised for resolving cryptic interactions and forecasting control outcomes. Future priorities include interdisciplinary integration, long-term monitoring, global comparisons, and addressing climate-driven uncertainties. This review aims to advance both theoretical invasion ecology and the development of sustainable, evidence-based biological control strategies.