Fuzzy Associational Rules and reasoning logic in computer vision models
Elouanes Khelifi, Usef Faghihi, Thierno Aliou Ba, Cyrus Kalantarpour, François Meunier · 2023
Many researchers aim towards solving human tasks with machine learning algorithms. In this paper we equipped YOLOV-5 with fuzzy logic rules that can be used to solve reasoning problems in classification (such as bottom-up classification) and linguistic feedback. Our method uses fuzzy associational rules that allow YOLOV-5 to deduce knowledge from a partial view of an object and can also be used to evaluate loss from fuzzy evaluations.