How useful is active learning for image-based plant phenotyping?

Abstract

Deep learning models have been successfully deployed for a diverse array of image-based plant phenotyping applications including disease detection and classification. However, successful deployment of supervised deep learning models requires large amount of labeled data, which is a significant challenge in plant sciences (and most biological) domain due to the inherent complexities. Specifically, data annotation is costly, laborious, time consuming and needs domain expertise for phenotyping tasks, especially for diseases. To overcome this challenge, active learning algorithms have been proposed to reduce the amount of labeling needed by deep learning models to achieve good predictive performance. Active learning methods work by adaptively suggesting samples to annotate using an acquisition function to achieve maximum (classification) performance under a fixed labeling budget. We report the performance of four different active learning methods, (1) Deep Bayesian Active Learning (DBAL), (2) Entropy, (3) Least Confidence, and (4) core-set, with conventional random sampling-based annotation for two vastly different image-based classification datasets. The first image dataset consists of soybean [Glycine max L. (Merr.)] leaves belonging to eight different soybean stresses and a healthy class, and the second consists of nine different weed species from the field. For a fixed labeling budget, we observed that the classification performance of deep learning models using active learning based acquisition strategies is better than random sampling-based acquisition for both datasets. The integration of active learning strategies for data annotation can help mitigate labelling challenges in the plant sciences applications particularly where resources dedicated to annotations are limited.