<link rel="stylesheet" href="styles.f3b1fba60ec7970c.css">

A Mobile-Based Deep Learning Model for Cassava Disease Diagnosis

dc.contributor.authorAmanda Ramcharan
dc.contributor.authorPeter McCloskey
dc.contributor.authorKelsee Baranowski
dc.contributor.authorNeema Mbilinyi
dc.contributor.authorLatifa Mrisho
dc.contributor.authorMathias Ndalahwa
dc.contributor.authorJames Legg
dc.contributor.authorDavid P. Hughes
dc.date.accessioned2026-11-18T07:11:34Z
dc.date.issued2019
dc.descriptionOfficial PMC article XML states CC BY 4.0 and permits use, sharing, adaptation, distribution and reproduction with attribution. Rights classification: FULL_TEXT_OK. Licence verified from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6436463/ (checked 2026-11-18). Source: International Institute for Tropical Agriculture, Dar el Salaam, Tanzania African relevance: Sector: Agriculture. Mobile CNN for cassava disease diagnosis tested on 720 leaflets in a Tanzanian field, with IITA co-authors. Official PMC article XML states CC BY 4.0 and permits use, sharing, adaptation, distribution and reproduction with attribution.
dc.description.abstractConvolutional 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.
dc.format.mimetypeapplication/pdf
dc.identifier.doi10.3389/fpls.2019.00272
dc.identifier.urihttps://doi.org/10.3389/fpls.2019.00272
dc.identifier.urihttps://library.dare.co.zw/handle/123456789/5102
dc.language.isoen
dc.publisherFrontiers Media SA
dc.rightsCC BY 4.0. Licence evidence: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6436463/
dc.rights.holderNamed authors
dc.rights.licenseCC BY 4.0
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.sourceInternational Institute for Tropical Agriculture, Dar el Salaam, Tanzania
dc.source.urihttps://doi.org/10.3389/fpls.2019.00272
dc.subjectAgriculture
dc.subjectcassava disease detection
dc.subjectdeep learning
dc.subjectconvolutional neural networks
dc.subjectmobile plant disease diagnostics
dc.subjectobject detection
dc.subjectFrontiers in Plant Science
dc.titleA Mobile-Based Deep Learning Model for Cassava Disease Diagnosis
dc.typejournal article

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
applied_agriculture_10_3389_fpls_2019_00272.pdf
Size:
5.29 MB
Format:
Adobe Portable Document Format