AXHU Students’ New Field Research

AXHU Students’ New Field Research

Mar 20, 2024, International Office

The AXHU postgraduate student HU Yating, C’24, also member of AXHU Research Group of AI-powered YOLO Detector, chaired and supervised by the presidential professor WANG Qijin, wrote the paper—MACNet: A More Accurate and Convenient Pest Detection Network, which was accepted and published on the world-leading journal Electronics (SCI, JCR2). To read more: https://doi.org/10.3390/electronics13061068

For collecting the big data of agricultural pests and diseases of varied types, small bodies, and high similarities, this research has been based on the algorithm of YOLOv8, which, via the technologies of CARMF (by reshaping and regrouping) and Distribution Convolution (DSConv), realizes a multiple of element-wise perceptive operations to critically solve the existing questions and problems in this field.

The CARMF is for mining much more featured deposit information and maximizing its extraction and testing ability of the update model accessible on the farming land, while the DSConv, on the other side, can reduce the size itself in some ways.

Empirically, the improved algorithm of MACNet can help obtain the testing performance of excellence regarding the data array of Pest24, which is already showing the big potential and real value at some complicated situations.

Over the past years, AXHU has been committed itself to improving the innovative and creative capacity of the undergraduate and graduate students, to grow their interests of research and practice, which is bringing out many outcomes and ideas of impact at AXHU community.

And the future research team will highlight the deep learning and research of how to combine with the Effective Receptive Field Theory and the Multiple-modality in the industrial areas, such as medical treatment, healthcare, agricultural pest monitoring, and more.