Method / resource 135,447 claims

Claim · model-written summary

GL-MobFormer achieved a classification accuracy of 95.13% on the tea leaf image dataset.

Taxon

teat0

Outcome

classification accuracye0

Value

95.13%e0

Verbatim

t0

“tea leaf phytopathogen diagnosis”

e0

“GL-MobFormer achieved a classification accuracy of 95.13%”

Claim · model-written summary

GL-MobFormer reached a Matthews Correlation Coefficient of 0.9417.

Taxon

teat0

Outcome

Matthews Correlation Coefficient (MCC)e0

Value

0.9417e0

Verbatim

t0

“tea leaf phytopathogen diagnosis”

e0

“a Matthews Correlation Coefficient (MCC) of 0.9417”

Claim · model-written summary

GL-MobFormer maintained its performance with a low computational footprint of 0.33 G FLOPs.

Taxon

teat0

Outcome

computational footprinte0

Value

0.33 G FLOPse0

Verbatim

t0

“tea leaf phytopathogen diagnosis”

e0

“a low computational footprint of merely 0.33 G FLOPs”

Claim · model-written summary

Masking only the top 5% of critical activation regions reduced classification confidence by an average of 70.61%, indicating the model relies on lesion features rather than background.

Taxon

teat0

Treatment

masking only the top 5% of critical activation regionse0

Outcome

classification confidencee0

Direction

reductione0

Value

70.61%e0

Verbatim

t0

“tea leaf phytopathogen diagnosis”

e0

“systematically masking only the top 5% of critical activation regions led to an average reduction of 70.61% in classification confidence”

Claim · model-written summary

The hyperspectral imaging plus improved YOLOv5 pipeline detected early latent bruises on mangoes with mAP_0.5 of 98.2%, with balanced precision and recall and fast inference at 320×320 resolution.

Taxon

mangoesMangifera indicat0

Outcome

mAP_0.5e1

Value

98.2%e1

Verbatim

t0

“mangoes”

e1

“The resulting pipeline detected early latent bruises with high accuracy, achieving mAP_0.5 = 98.2% with balanced precision/recall and fast inference at 320×320 resolution”

Claim · model-written summary

Among pre-trained and custom CNN architectures tested for orange disease classification, the custom scratch-built CNN achieved the best results.

Taxon

oranget0

Outcome

best resultse0

Direction

beste0

Comparator

pre-trainede0

Verbatim

t0

“orange fruit diseases”

e0

“both pre-trained and custom CNN architectures were tested, with the custom, scratch-built CNN achieving the best results”

Claim · model-written summary

The proposed CNN model reached an accuracy of 97% in classifying orange fruit diseases.

Taxon

oranget0

Outcome

accuracye0

Value

97%e0

Verbatim

t0

“orange fruit diseases”

e0

“The proposed model reached an accuracy of 97%”

Claim · model-written summary

Multiple regression models incorporating soil nitrogen and root variables explained more than 50% of the variance in summer flush phenotype.

Taxon

citrusCitrust0

Outcome

variance in SF phenotypee1

Value

>50 %e1

Verbatim

t0

“citrus”

e1

“Multiple regression models incorporating these variables explained >50 % of the variance in SF phenotype”

Claim · model-written summary

On a real-world dataset of 1,216 images across five cultivars, the detector improved AP metrics and error counts over the YOLOv8n baseline at comparable runtime.

Taxon

grapevineVitist0

Outcome

AP metrics and error countse0

Direction

improvementse0

Comparator

YOLOv8n baselinee0

Verbatim

t0

“grapevine”

e0

“Experiments on a real-world dataset of 1,216 images across five cultivars show consistent improvements over the YOLOv8n baseline in AP metrics and error counts at comparable runtime”

Claim · model-written summary

Hybrid-Linear post-processing reached an inference throughput of about 32 FPS on a server-grade GPU.

Taxon

grapevineVitist0

Outcome

inference throughpute0

Value

≈32 FPSe0

Verbatim

t0

“grapevine”

e0

“Hybrid-Linear achieves an inference throughput of ≈32 FPS on a server-grade GPU under our evaluation protocol”