A Mobile-Based Deep Learning Model for Cassava Disease Diagnosis.

Mathias Ndalahwa,Amanda Ramcharan,Latifa Mrisho,Neema Mbilinyi,David P Hughes,Peter Mccloskey,James Legg,Kelsee Baranowski

doi:10.3389/fpls.2019.00272

Abstract

Convolutional neural network (CNN) models have the potential to improve plant disease phenotyping where the standard approach is visual diagnostics requiring specialized training. In scenarios where a CNN is deployed on mobile devices, models are presented with new challenges due to lighting and orientation. It is essential for model assessment to be conducted in real world conditions if such models are to be reliably integrated with computer vision products for plant disease phenotyping. We train a CNN object detection model to identify foliar symptoms of diseases in cassava (Manihot esculenta Crantz). We then deploy the model in a mobile app and test its performance on mobile images and video of 720 diseased leaflets in an agricultural field in Tanzania. Within each disease category we test two levels of severity of symptoms-mild and pronounced, to assess the model performance for early detection of symptoms. In both severities we see a decrease in performance for real world images and video as measured with the F-1 score. The F-1 score dropped by 32% for pronounced symptoms in real world images (the closest data to the training data) due to a decrease in model recall. If the potential of mobile CNN models are to be realized our data suggest it is crucial to consider tuning recall in order to achieve the desired performance in real world settings. In addition, the varied performance related to different input data (image or video) is an important consideration for design in real world applications.

Highlights

Conventional plant disease diagnosis by human experts is inherently subjective and limited to regions that can support the required human infrastructure (Bock et al, 2010)
In this study we evaluate the performance of a Convolutional neural network (CNN) model deployed offline in real time on a mobile device to detect foliar symptoms of cassava pests and diseases
Using the singleshot detector model, a CNN architecture optimized for mobile devices, we assess the performance of the model to detect pronounced and mild symptoms of 3 disease classes

Summary

Introduction

Conventional plant disease diagnosis by human experts is inherently subjective and limited to regions that can support the required human infrastructure (Bock et al, 2010). Computer vision algorithms show promise to transform this field with the landmark result of a deep convolutional neural network (CNN) winning the Imagenet competition to classify over 1 million images from 1,000 categories, almost halving the error rates of its competition (LeCun et al, 2015). This success brought about a revolution in computer vision with CNN models dominating the approach for a variety of classification and detection tasks. Deploying on mobile devices would be beneficial in democratizing access to algorithms while maintaining user privacy by running inference offline

Methods

Results

Conclusion