Abstract
Wheat leaf rust is a significant disease affecting wheat crops, leading to substantial yield losses. Early and accurate detection of this disease is crucial for timely intervention and management. This study explores the application of small CNN architectures, specifically MobileNetV2 and ShuffleNet, for the detection of wheat leaf rust. A dataset comprising 491 healthy leaf samples and 368 leaf rust samples was utilized to evaluate the performance of these models. MobileNetV2 achieved an Acc. of 98.65%, with a Sens. of 100%, Spec. of 97.30%, and a precision of 97.37%. In comparison, ShuffleNet attained an Acc. of 94.59%, Sens. of 100%, Spec. of 89.19%, and a precision of 90.24%. The respective F1 scores for MobileNetV2 and ShuffleNet were 98.67% and 94.87%, while the MCC values were 0.9733 and 0.8971, and the Kappa coefficients were 0.9730 and 0.8919. The computational times for MobileNetV2 and ShuffleNet were 14.2697 seconds and 13.5902 seconds, respectively. The confusion matrices for both models further illustrate their classification performance. The results indicate that both MobileNetV2 and ShuffleNet are effective in detecting wheat leaf rust, with MobileNetV2 demonstrating superior overall performance. These findings suggest the potential for deploying these lightweight CNN models in mobile applications for real-time monitoring and management of wheat leaf rust.
Abstract from the paper’s OpenAlex record, shown for indexing and research. Copyright remains with the authors or the publisher. Read at the source ↗
The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.
Claims 17
Claim · model-written summary
MobileNetV2 detected wheat leaf rust with 98.65% accuracy.
- Taxon
- Wheat
Triticum aestivumt0 - Outcome
- Acc.e0
- Value
- 98.65%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“MobileNetV2 achieved an Acc. of 98.65%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 had 100% sensitivity.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Sens.e0
- Value
- 100%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“with a Sens. of 100%, Spec. of 97.30%, and a precision of 97.37%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 had 97.30% specificity.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Spec.e0
- Value
- 97.30%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Spec. of 97.30%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 had 97.37% precision.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- precisione0
- Value
- 97.37%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“precision of 97.37%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet detected wheat leaf rust with 94.59% accuracy.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Acc.e0
- Value
- 94.59%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“ShuffleNet attained an Acc. of 94.59%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet had 100% sensitivity.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Sens.e0
- Value
- 100%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“In comparison, ShuffleNet attained an Acc. of 94.59%, Sens. of 100%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet had 89.19% specificity.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Spec.e0
- Value
- 89.19%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Spec. of 89.19%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet had 90.24% precision.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- precisione0
- Value
- 90.24%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“precision of 90.24%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 F1 score was 98.67%.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- F1 scorese0
- Value
- 98.67%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The respective F1 scores for MobileNetV2 and ShuffleNet were 98.67% and 94.87%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet F1 score was 94.87%.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- F1 scorese0
- Value
- 94.87%e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“F1 scores for MobileNetV2 and ShuffleNet were 98.67% and 94.87%”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 MCC was 0.9733.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- MCC valuese0
- Value
- 0.9733e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“MCC values were 0.9733 and 0.8971”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet MCC was 0.8971.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- MCC valuese0
- Value
- 0.8971e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“MCC values were 0.9733 and 0.8971”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 Kappa coefficient was 0.9730.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Kappa coefficientse0
- Value
- 0.9730e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Kappa coefficients were 0.9730 and 0.8919”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet Kappa coefficient was 0.8919.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- Kappa coefficientse0
- Value
- 0.8919e0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Kappa coefficients were 0.9730 and 0.8919”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 computation took 14.2697 seconds.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- computational timese0
- Value
- 14.2697 secondse0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The computational times for MobileNetV2 and ShuffleNet were 14.2697 seconds and 13.5902 seconds, respectively”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
ShuffleNet computation took 13.5902 seconds.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- computational timese0
- Value
- 13.5902 secondse0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The computational times for MobileNetV2 and ShuffleNet were 14.2697 seconds and 13.5902 seconds, respectively”
t0“Wheat Leaf Rust Detection”
Claim · model-written summary
MobileNetV2 had superior overall wheat leaf rust detection performance.
- Taxon
- WheatTriticum aestivumt0
- Outcome
- overall performancee0
- Direction
- superiore0
- Comparator
- ShuffleNete1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“MobileNetV2 demonstrating superior overall performance”
e1“both MobileNetV2 and ShuffleNet are effective in detecting wheat leaf rust”
t0“Wheat Leaf Rust Detection”