Try:droughtFusariumnitrogen fertilizerArabidopsisseed germination

“decrease in A”

126 matches

Claim types

Direction of the findings

plus 19 non-evaluative

Over time

1950 ’60 ’70 ’80 ’90 2000 ’10 ’20

Refine ·

More filters

Clear filters

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

CPD-YOLO generalization showed a 0.55 % mAP decrease.

Taxon: cottonGossypium hirsutumt0
Outcome: mAPe0
Direction: decreasee0
Value: 0.55 %e0

Not reportedComparator

Verbatim

e0
“the CPD-YOLO model had superior generalization capability, with a 1.30 % decrease in F1-score and a 0.55 % decrease in mAP.”
✓ located verbatim in the source · abstract, characters 1287–1410

t0
“cotton pests and diseases”
✓ located verbatim in the source · title, characters 48–73

Title

CPD-YOLO: A cross-platform detection method for cotton pests and diseases using UAV and smartphone imaging

Guowei Yu; Benxue Ma; Ruoyu Zhang; et al. Industrial Crops and Products 2025. doi:10.1016/j.indcrop.2025.121515

Other claims from this paper (7)

Method / resource The improvement strategy improved multi-scale object detection accuracy in cross-platform scenarios.
Method / resource CPD-YOLO increased F1-score by 7.44 % versus the original model.
Method / resource CPD-YOLO increased mAP by 7.08 % versus the original model.
Method / resource CPD-YOLO outperformed typical detectors with an F1-score of 88.86 %.
Method / resource CPD-YOLO outperformed typical detectors with mAP of 90.42 %.
Method / resource CPD-YOLO generalization showed a 1.30 % F1-score decrease.
Method / resource A mobile application named Doctor Cotton using CPD-YOLO was developed.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

CPD-YOLO generalization showed a 1.30 % F1-score decrease.

Taxon: cottonGossypium hirsutumt0
Outcome: F1-scoree0
Direction: decreasee0
Value: 1.30 %e0

Not reportedComparator

Verbatim

e0
“the CPD-YOLO model had superior generalization capability, with a 1.30 % decrease in F1-score and a 0.55 % decrease in mAP.”
✓ located verbatim in the source · abstract, characters 1287–1410

t0
“cotton pests and diseases”
✓ located verbatim in the source · title, characters 48–73

Title

CPD-YOLO: A cross-platform detection method for cotton pests and diseases using UAV and smartphone imaging

Guowei Yu; Benxue Ma; Ruoyu Zhang; et al. Industrial Crops and Products 2025. doi:10.1016/j.indcrop.2025.121515

Other claims from this paper (7)

Method / resource The improvement strategy improved multi-scale object detection accuracy in cross-platform scenarios.
Method / resource CPD-YOLO increased F1-score by 7.44 % versus the original model.
Method / resource CPD-YOLO increased mAP by 7.08 % versus the original model.
Method / resource CPD-YOLO outperformed typical detectors with an F1-score of 88.86 %.
Method / resource CPD-YOLO outperformed typical detectors with mAP of 90.42 %.
Method / resource CPD-YOLO generalization showed a 0.55 % mAP decrease.
Method / resource A mobile application named Doctor Cotton using CPD-YOLO was developed.

Method / resourceobservedoriginal resultnegativemeasured

Claim · model-written summary

Models with one and four augmentation techniques showed a decrease in performance.

Taxon: Tea Leaft0
Outcome: performancee0
Direction: decreasee0
Comparator: one and four augmentation techniquese0

Verbatim

t0
“Tea Leaf Disease Classification”
✓ located verbatim in the source · title, characters 83–114

e0
“models with one and four augmentation techniques experience a decrease in performance”
✓ located verbatim in the source · abstract, characters 889–974

Title

Comparative Analysis of Data Augmentation Techniques on MobileNetV2 Performance in Tea Leaf Disease Classification

Fitri Mintarsih; Dewi Khairani; M. Farhan 2024 12th International Conference on Cyber and IT Service Management (CITSM) 2024. doi:10.1109/citsm64103.2024.10775557

Other claims from this paper (5)

Method / resource Variation in combinations of data augmentation techniques in the MobileNetV2 model did not affect changes in accuracy, loss, precision, recall or F1-score in tea leaf disease classification.
Method / resource The use of augmentation techniques significantly affected the model performance.
Method / resource Models with three and five augmentation techniques showed very good results.
Method / resource Model number 32, using flip, rotation, translation, zoom and contrast augmentation, had the highest accuracy of 0.9867.
Method / resource Model number 32 had a low loss value of 0.0603.

Method / resourceobservedoriginal resultnegativemeasured

Claim · model-written summary

All four models had a slight accuracy decrease on workstations and mobile devices.

Taxon: Plantt0
Outcome: accuracye1
Direction: decreasee0
Comparator: workstation and mobile devicese0

Verbatim

e0
“the four models have a slight accuracy decrease on both workstation and mobile devices”
✓ located verbatim in the source · abstract, characters 765–851

e1
“Mason model has a highest accuracy of 90.54%”
✓ located verbatim in the source · abstract, characters 567–611

t0
“Plant Diseases”
✓ located verbatim in the source · title, characters 40–54

Title

Evaluation of CNN Models in Identifying Plant Diseases on a Mobile Device

Teddy Aristan; Gede Putra Kusuma Revue d intelligence artificielle 2023. doi:10.18280/ria.370221

Other claims from this paper (7)

Method / resource Mason reached the highest accuracy, 90.54%, for plant disease classification.
Method / resource Mason had the smallest output file, 0.85MB.
Method / resource MobileNetV3 achieved 88.83% accuracy and a 16.85MB file.
Method / resource EfficientNetB0 achieved 88.75% accuracy and a 16.08MB file.
Method / resource ShuffleNetV2 achieved 83.52% accuracy and a 15.89MB file.
Method / resource MobileNetV3 consumed fewer resources overall than the other models.
Recommendation The authors recommend Mason for mobile devices based on accuracy and low resource consumption.

Method / resourceobservedoriginal resultnegativemeasured

Claim · model-written summary

A qPCR detection method confirmed a twofold decrease of health indicators in beet clamps.

Taxon: beetBeta vulgarist0
Outcome: health indicatorse0
Direction: decreasee0
Value: twofolde0
Qualifier: qPCR-based early detection techniquee0

Verbatim

e0
“a qPCR-based early detection technique was developed and confirmed a twofold decrease of health indicators and an up to 10,000-fold increase of disease indicators in beet clamps”
✓ located verbatim in the source · abstract, characters 1485–1662

t0
“sugar beets”
✓ located verbatim in the source · title, characters 85–96

Title

Microbiome-driven identification of microbial indicators for postharvest diseases of sugar beets

Peter Kusstatscher; Christin Zachow; Karsten Harms; et al. Microbiome 2019. doi:10.1186/s40168-019-0728-0

Other claims from this paper (8)

Observation Stored beet microbiomes differed regionally across clamps in Austria and Germany.
Observation Microbial diversity significantly decreased in decaying sugar beets.
Observation The overall microbiome composition shifted in decaying sugar beets.
Observation Candida, Penicillium, and Lactobacillus were main disease indicators in decaying sugar beets.
Observation Plectosphaerella and Vishniacozyma reflected the microbiome of healthy beets.
Association Higher microbial diversity reflected healthy beet microbiomes.
Method / resource A qPCR detection method confirmed up to a 10,000-fold increase of disease indicators in beet clamps.
Method / resource Sugar content analyses verified the microbial indicator findings.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

The improved model achieved a 28.0% decrease in computational volume relative to YOLOv8n.

Taxon: ApplesMalus domesticat0
Outcome: computational volumee0
Direction: decreasee0
Value: 28.0%e0
Comparator: YOLOv8n modele0

Verbatim

e0
“Experimental results indicate that the improved model surpasses the YOLOv8n model, achieving a 2.7% increase in average accuracy, a 24.3% reduction in parameters, a 28.0% decrease in computational volume, and an 8.5% improvement in inference speed.”
✓ located verbatim in the source · abstract, characters 1098–1346

t0
“Apples”
✓ located verbatim in the source · title, characters 50–56

Title

Lightweight Non-Destructive Detection of Diseased Apples Based on Structural Re-Parameterization Technique

Bo Han; Ziao Lu; Luan Dong; et al. Applied Sciences 2024. doi:10.3390/app14051907

Other claims from this paper (3)

Method / resource The improved model achieved a 2.7% increase in average accuracy relative to YOLOv8n.
Method / resource The improved model achieved a 24.3% reduction in parameters relative to YOLOv8n.
Method / resource The improved model achieved a 8.5% improvement in inference speed relative to YOLOv8n.

Method / resourceobservedoriginal resultnegativemeasured

Claim · model-written summary

Low-resolution testing gave a 2% decrease in metrics compared with high-resolution scenarios.

Taxon: riceOryza sativae0
Outcome: metricse2
Direction: decreasee2
Value: 2%e2
Comparator: high-resolution scenariose2

Verbatim

e0
“Detecting diseases on rice leaves plays a crucial role”
✓ located verbatim in the source · abstract, characters 0–54

e2
“Notably, Tiny-CNN trained on high-resolution images excels, achieving over 98% across all metrics when tested on low-resolution images, with a 2% decrease compared to high-resolution scenarios”
✓ located verbatim in the source · abstract, characters 758–950

Title

Tiny-CNN: Structuring Convolutional Neural Networks for Accurate Classification of Rice Leaf Diseases in Resource-Constrained Environments

Khang Quoc Nguyen; Huy Cao Gia Nguyen; Trinh Ngo Diem Le; et al. Proceedings of the 2024 9th International Conference on Intelligent Information Technology 2024. doi:10.1145/3654522.3654570

Other claims from this paper (4)

Method / resource Tiny-CNN accuracy for rice leaf disease exceeded 91.8%.
Method / resource Tiny-CNN F1-score for rice leaf disease exceeded 92%.
Method / resource Tiny-CNN trained on high-resolution images achieved over 98% across all metrics when tested on low-resolution images.
Method / resource Tiny-CNN trained on limited low-resolution data maintained results above 95%.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

The model reduced FLOPs by 50% relative to YOLOv5s.

Taxon: riceOryza sativae0
Outcome: FLOPse0
Direction: decreasee0
Value: 50%e0
Comparator: YOLOv5se0

Verbatim

e0
“The experimental results showed that the designed model had a 49% decrease in the number of model parameters and a 50% decrease in FLOPs without a decrease in precision on self-built rice blast dataset, compared with the YOLOv5s”
✓ located verbatim in the source · abstract, characters 688–916

Title

A LIGHTWEIGHT IMPROVED YOLOv5s MODEL-BASED RICE BLAST DETECTION METHOD AND MOBILE DEPLOYMENT

Fankai MENG; Congkuan YAN; Yuqing YANG; et al. INMATEH Agricultural Engineering 2024. doi:10.35633/inmateh-74-68

Other claims from this paper (3)

Method / resource The model reduced parameter count by 49% relative to YOLOv5s.
Method / resource The model maintained precision relative to YOLOv5s.
Method / resource RiceBlastDetector accurately detected disease spots at different scales in field experiments.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Decrease in height growth from fungal attack satisfactorily assessed needle cast resistance.

Taxon: Scots pinePinus sylvestrist0
Interactor: LophodermiumLophodermiumt0pathogen
Outcome: resistancee0
Comparator: decrease in height growthe0

Verbatim

e0
“Determination of the decrease in height growth caused by the fungal attack was a satisfactory method for assessing the resistance”
✓ located verbatim in the source · abstract, characters 158–287

t0
“Testing Scots pine for resistance to Lophodermium needle cast”
✓ located verbatim in the source · title, characters 0–61

Title

Testing Scots pine for resistance to Lophodermium needle cast

O. Martinsson Epsilon Open Archive (Sveriges lantbruksuniversitet biblioteket (Swedish University of Agricultural Sciences)) 1979

Other claims from this paper (1)

Observation Height growth of the most susceptible pine progenies was about 30% of that of the most resistant progenies.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

The model reduced parameter count by 49% relative to YOLOv5s.

Taxon: riceOryza sativae0
Outcome: number of model parameterse0
Direction: decreasee0
Value: 49%e0
Comparator: YOLOv5se0

Verbatim

e0
“The experimental results showed that the designed model had a 49% decrease in the number of model parameters and a 50% decrease in FLOPs without a decrease in precision on self-built rice blast dataset, compared with the YOLOv5s”
✓ located verbatim in the source · abstract, characters 688–916

Title

A LIGHTWEIGHT IMPROVED YOLOv5s MODEL-BASED RICE BLAST DETECTION METHOD AND MOBILE DEPLOYMENT

Fankai MENG; Congkuan YAN; Yuqing YANG; et al. INMATEH Agricultural Engineering 2024. doi:10.35633/inmateh-74-68

Other claims from this paper (3)

Method / resource The model reduced FLOPs by 50% relative to YOLOv5s.
Method / resource The model maintained precision relative to YOLOv5s.
Method / resource RiceBlastDetector accurately detected disease spots at different scales in field experiments.

Lexical search (BM25). All words must match; if nothing does, any word. Organism names also match every spelling of the same organism. No semantic search yet.

“decrease in A”

Claim types

Direction of the findings

Over time

Claim type

Research area

Polarity

Statistics