EasyDAM_V2: Efficient Data Labeling Method for Multishape, Cross-Species Fruit Detection
作者机构:Information DepartmentBeijing University of TechnologyBeijing 100022China Graduate School of Agricultural and Life SciencesThe University of TokyoTokyo 188-0002Japan
出 版 物:《Plant Phenomics》 (植物表型组学(英文))
年 卷 期:2022年第4卷第1期
页 面:94-109页
核心收录:
学科分类:0710[理学-生物学] 08[工学] 09[农学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 0901[农学-作物学] 0902[农学-园艺学] 090201[农学-果树学] 0812[工学-计算机科学与技术(可授工学、理学学位)]
基 金:This study was partially supported by the National Natural Science Foundation of China(NSFC)Program U19A2061,International Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences(CAASTIP) Japan Science and Technology Agency(JST)AIP Accel-eration Research JPMJCR21U3.
摘 要:In modern smart orchards,fruit detection models based on deep learning require expensive dataset labeling work to support the construction of detection models,resulting in high model application costs.Our previous work combined generative adversarial networks(GANs)and pseudolabeling methods to transfer labels from one specie to another to save labeling costs.However,only the color and texture features of images can be migrated,which still needs improvement in the accuracy of the data labeling.