U-FedTomAtt: Ultra-lightweight Federated Learning with Attention for Tomato Disease Recognition
U-FedTomAtt:轻量级联邦学习与注意力机制用于番茄疾病识别
机构 * Department of Computer Engineering, University of Peradeniya(珀德尼亚大学计算机工程系) ; Faculty of Science and Technology, Charles Darwin University(查尔斯达尔文大学科学与技术学院) ; Canadian Institute for Cybersecurity, University of New Brunswick(新不伦瑞克大学加拿大网络安全研究所) ; School of Mechanical and Electrical Engineering, Guangzhou University(广州大学机械与电子工程学院)
专题命中 医学数据与评测 :diagnosis(abstract);分类 cs.LG、q-bio
AI总结 U-FedTomAtt提出一种轻量级联邦学习框架,结合注意力机制,用于在资源受限环境下识别番茄疾病,实现了高准确率和低计算开销。
Comments 10 pages and 4 figures