用于稳健昆虫分类的越界检测算法（英文版）

英文标题：Out-of-distribution detection algorithms for robust insect classification中文摘要：本文研究了在农业领域应用昆虫识别和分类模型的难点，提出使用基于深度学习的算法进行准确性检测，并评估了目前基于最大软最大概率、马氏距离以及基于能量的三种 OOD 算法对昆虫识别分类器性能的影响。英文摘要：Deep learning-based approaches have produced models with good insectclassification accuracy; Most of these models are conducive for application incontrolled environmental conditions. One of the primary emphasis of researchersis to implement identification and classification models in the realagriculture fields, wh