Mr. Keyi Chen | Jihua Laboratory | China
Keyi Chen is a dedicated research engineer at Jihua Laboratory, Foshan, Guangdong Province, China. He obtained his MSc in Crop Informatics from Huazhong Agricultural University , where he built a strong foundation in computational modeling and artificial intelligence applications. His research primarily focuses on deep learning algorithms, particularly their integration into computer-based recognition systems and intelligent environmental analysis. He has completed three research projects and participated in one industry consultancy project, demonstrating both academic and applied innovation. His current research explores AI-driven recognition of marine microalgae, an essential area for assessing aquatic ecological health. In this domain, Chen developed a ResNeXt-50-based multi-expert network with an exponential feature compression mechanism that effectively mitigates class imbalance issues. Evaluated on the WHIO-Plankton dataset, his model achieved a state-of-the-art performance with an average precision and average recall , outperforming existing baselines. The system’s low inference latency demonstrates high real-time feasibility. His contributions provide a robust framework for marine microalgae recognition, supporting environmental monitoring and marine life science research. With Citations by 38 documents, 3 publications, and an h-index of 2, Chen has established himself as a rising researcher in applied AI and computational biology. His ongoing innovations signify impactful potential in environmental intelligence, sustainable technology, and bioinformatics applications.
Featured Publications
Chen, K., Cui, S., Zhong, J., & Wang, Q. (2025). MicroalgaeNet: Enhancing recognition of long-tailed marine microalgae images through multi-expert networks and feature compression. Algal Research, 92, 104333. https://doi.org/10.1016/j.algal.2025.104333
Song, P., Chen, K., Zhu, L., Yang, M., Ji, C., Xiao, A., Jia, H., Zhang, J., & Yang, W. (2022). An improved cascade R-CNN and RGB-D camera-based method for dynamic cotton top bud recognition and localization in the field. Computers and Electronics in Agriculture, 202, 107442. https://doi.org/10.1016/j.compag.2022.107442