Few-shot learning for biotic stress classification of coffee leaves

In the last few years, deep neural networks have achieved promising results in several fields. However, one of the main limitations of these methods is the need for large-scale datasets to properly generalize. Few-shot learning methods emerged as an attempt to solve this shortcoming. Among the few-s...

Full description

Bibliographic Details
Main Authors: Lucas M. Tassis, Renato A. Krohling
Format: Article
Language:English
Published: KeAi Communications Co., Ltd. 2022-01-01
Series:Artificial Intelligence in Agriculture
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2589721722000046