Try:droughtFusariumnitrogen fertilizerArabidopsisseed germination

“inference time”

1,108 matches

Claim types

Direction of the findings

plus 291 non-evaluative

Over time

1970 ’80 ’90 2000 ’10 ’20

Refine

More filters

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Inference time was 18.34 s.

Outcome: inference timee0
Direction: lowe0
Value: 18.34 se0

Not reportedTaxon

Verbatim

e0
“low inference time of 18.34 s”
✓ located verbatim in the source · abstract, characters 1269–1298

Title

DeepGreen: a real-time deep learning system for smart agriculture monitoring

Ajit Singh Rathor; Sushabhan Choudhury; Abhinav Sharma; et al. Scientific Reports 2026. doi:10.1038/s41598-026-45395-3

Other claims from this paper (10)

Method / resource Conv-7 DCNN classifies tomato leaf diseases within a fifteen-category system.
Method / resource Conv-7 DCNN classifies potato leaf diseases within a fifteen-category system.
Method / resource Conv-7 DCNN classifies pepper-bell leaf diseases within a fifteen-category system.
Method / resource The model outperformed several pretrained and trending deep learning models.
Method / resource Classification accuracy was 99.18%.
Method / resource Average precision was 99.17%.
Method / resource Area under the curve was 1.
Method / resource Conv-7 DCNN achieved FPS of 112.49.
Method / resource GFLOPS was 13.98.
Recommendation The model is suitable for real-time smart agriculture applications.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

Inference time was on a millisecond scale.

Taxon: CottonGossypium hirsutumt0
Outcome: inference timee0
Value: millisecond-scalee0

Verbatim

e0
“millisecond-scale inference time for real-time processing”
✓ located verbatim in the source · abstract, characters 763–820

t0
“Cotton”
✓ located verbatim in the source · title, characters 27–33

Title

Spatio-temporal Mapping of Cotton Blooms Appearance Using Deep Learning

Vaishnavi Thesma; Canicius Mwitta; Glen Rains; et al. IFAC-PapersOnLine 2022. doi:10.1016/j.ifacol.2022.11.111

Other claims from this paper (3)

Method / resource The best model counted cotton flowers with F1 score above 0.95.
Method / resource The model had a 98% true positive rate.
Method / resource False negative and false positive rates were both under 10%.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Int8 quantization reduced weed classifier inference time.

Taxon: weedt0
Outcome: inference timee0
Direction: reductionse0

Not reportedComparatorValue

Verbatim

e0
“staggering model size and inference time reductions while maintaining accuracy”
✓ located verbatim in the source · abstract, characters 789–867

t0
“Weed”
✓ located verbatim in the source · title, characters 16–20

Title

QWID: Quantized Weed Identification Deep Neural Network

Parikshit Singh Rathore Proceedings of the Conference on Robots and Vision 2024. doi:10.21428/d82e957c.0b4519c5

Other claims from this paper (1)

Method / resource Int8 quantization reduced weed classifier model size relative to fp32 versions.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

MRD-YOLO retained real-time GPU inference and exceptional CPU inference speed.

Outcome: inference capabilitye0
Setting: GPU environmentse0

Not reportedComparatorValue

Verbatim

e0
“it maintains real-time inference capability in GPU environments and demonstrates exceptional inference speed in CPU environments.”
✓ located verbatim in the source · abstract, characters 1175–1304

Title

Melon ripeness detection by an improved object detection algorithm for resource constrained environments

Xuebin Jing; Yuanhao Wang; Dongxi Li; et al. Plant Methods 2024. doi:10.1186/s13007-024-01259-3

Other claims from this paper (6)

Method / resource A greenhouse-sourced melon ripeness dataset was created with occlusion, overlap and lighting variation.
Method / resource MRD-YOLO achieved 97.4% mean Average Precision on the melon dataset.
Method / resource MRD-YOLO used 4.8 G FLOPs, 58.5% of the YOLOv8n baseline.
Method / resource MRD-YOLO used 2.06 M parameters, 68.4% of the YOLOv8n baseline.
Method / resource MRD-YOLO outperformed existing methods in balanced accuracy and computational efficiency.
Method / resource MRD-YOLO reached 85.9% mean Average Precision on two unseen Roboflow melon datasets.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

The model inference time was 31.4 ms.

Taxon: tomatoSolanum lycopersicumt0
Outcome: inference timee0
Value: 31.4 mse0

Verbatim

e0
“inference time of 31.4 ms”
✓ located verbatim in the source · abstract, characters 1979–2004

t0
“tomato”
✓ located verbatim in the source · title, characters 39–45

Title

A novel YOLOv11 framework for enhanced tomato disease detection

Entesar Hamed I. Eliwa; Tarek Abd El-Hafeez PeerJ Computer Science 2025. doi:10.7717/peerj-cs.3200

Other claims from this paper (6)

Method / resource Enhanced YOLOv11 achieved 99.93% overall accuracy for tomato disease classification.
Method / resource Many disease classes reached 100.00% precision, recall and F1 scores.
Method / resource The full AGMS-FF variant achieved 99.84% accuracy.
Method / resource The model required 126 min for training.
Method / resource The model used 3.2 GB of memory.
Method / resource The model throughput was 38.2 FPS.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

YOLOv8 maintained substantially lower inference time.

Outcome: inference timee0
Direction: lowere0
Comparator: lowere0

Verbatim

e0
“maintaining substantially lower inference time”
✓ located verbatim in the source · abstract, characters 1091–1137

Title

Comparative evaluation of deep learning models for plant disease classification with edge-aware performance analysis

Zirije Hasani; Jakup Fondaj; Dionesa Bytyqi; et al. PLoS ONE 2026. doi:10.1371/journal.pone.0349901

Other claims from this paper (7)

Method / resource DenseNet121 classified disease with 99.48% ± 0.12 accuracy.
Method / resource DenseNet121 had significantly higher inference latency.
Method / resource BNN achieved minimal latency.
Method / resource BNN classification performance fell to 88.31% ± 0.45.
Method / resource YOLOv8-cls achieved 99.64% ± 0.09 accuracy.
Method / resource YOLOv8-cls inference latency was 3.3 ms ± 0.2.
Method / resource YOLOv8 significantly outperformed ResNet50.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

Average inference time was 14.33 ms.

Taxon: Sugarcanet0
Outcome: average inference timee0
Value: 14.33 mse0

Verbatim

e0
“14.33 ms average inference time”
✓ located verbatim in the source · abstract, characters 1272–1303

t0
“sugarcane”
✓ located verbatim in the source · title, characters 10–19

Title

Real time sugarcane stem health detection with YOLOv8 tinier and RepNCSP ELAN

Pushpendra Kumar; Shraddha Arora; Shaveta Arora Scientific Reports 2025. doi:10.1038/s41598-025-27502-y

Other claims from this paper (8)

Method / resource The detection model uses 7.2 GFLOPs.
Method / resource The model has 2.2 million parameters.
Method / resource The model size is 4.6 MB in PyTorch.
Method / resource The model achieved 90.1% mAP@50.
Method / resource The model achieved 64.2% mAP@95.
Method / resource The model had lower complexity than earlier nano YOLO versions.
Method / resource Benchmarking found superior efficiency and competitive accuracy.
Method / resource The model was validated for deployment across edge AI platforms and formats.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Deployed on the Jetson Orin Nano without inference acceleration, TMSDDet achieved stable real-time performance with an inference time of 31.0 ms.

Taxon: TomatoSolanum lycopersicumt0
Outcome: inference timee0
Value: 31.0 mse0
Setting: Jetson Orin Nanoe0

Verbatim

t0
“Tomato Maturity and Defect Detection”
✓ located verbatim in the source · title, characters 33–69

e0
“the model was successfully deployed on the Jetson Orin Nano without inference acceleration, achieving stable real-time performance (inference time: 31.0 ms)”
✓ located verbatim in the source · abstract, characters 1584–1740

Title

TMSDDET: A Lightweight Model for Tomato Maturity and Defect Detection on Edge Devices

Zhiyong Ma; Huili Zhang; Yuliang Yun; et al. Applied Engineering in Agriculture 2026. doi:10.13031/aea.16497

Other claims from this paper (6)

Method / resource TMSDDet achieved 80.4% mAP@0.5:0.95 for tomato maturity and defect detection.
Method / resource TMSDDet has only 1.9M parameters.
Method / resource TMSDDet requires 3.9 GFLOPs.
Method / resource TMSDDet had an inference time of 6.2 ms per frame on an RTX 4060.
Method / resource Compared with mainstream models such as YOLOv6, YOLOv8 and RT-DETR, TMSDDet delivered competitive performance on the authors test set.
Method / resource TMSDDet has a compact model size of just 4.0 MB.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

Average inference time was under 40 ms.

Taxon: cornZea mays subsp. mayst0
Outcome: average inference timee0
Value: under 40 mse0

Verbatim

e0
“an average inference time of under 40 ms”
✓ located verbatim in the source · abstract, characters 1474–1514

t0
“corn”
✓ located verbatim in the source · title, characters 35–39

Title

IRG-ResNet: distillation model for corn disease recognition

Shaoqiu Zhu; Lujie Bai; Haitao Gao Complex & Intelligent Systems 2025. doi:10.1007/s40747-025-02203-1

Other claims from this paper (2)

Method / resource The student network achieved 98.25% corn disease recognition accuracy after distillation.
Method / resource The model parameter size was under 5 MB.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

YOLOv8nGP inference time was 2.2 ms.

Outcome: inference timee0
Value: 2.2 mse0

Not reportedTaxon

Verbatim

e0
“the inference time was 2.2 ms”
✓ located verbatim in the source · abstract, characters 765–794

t0
“Soybean”
✓ located verbatim in the source · title, characters 20–27

Title

Weed Recognition at Soybean Seedling Stage Based on YOLOV8nGP + NExG Algorithm

Tao Sun; Longfei Cui; Lixuan Zong; et al. Agronomy 2024. doi:10.3390/agronomy14040657

Other claims from this paper (5)

Method / resource YOLOv8nGP improved soybean recognition precision by 1.1% over YOLOv8n.
Method / resource YOLOv8nGP improved soybean recognition recall by 1.1% over YOLOv8n.
Method / resource YOLOv8nGP improved soybean recognition F1 by 1.1% over YOLOv8n.
Method / resource YOLOv8nGP reduced model size by 3.6 mb.
Method / resource The segmentation method reduced dependence on weed dataset diversity.

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.

“inference time”

Claim types

Direction of the findings

Over time

Claim type

Research area

Polarity

Statistics