2022
DOI: 10.1016/j.eja.2022.126607
|View full text |Cite
|
Sign up to set email alerts
|

Accurate modeling of vertical leaf nitrogen distribution in summer maize using in situ leaf spectroscopy via CWT and PLS-based approaches

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
8
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
8

Relationship

0
8

Authors

Journals

citations
Cited by 16 publications
(8 citation statements)
references
References 76 publications
0
8
0
Order By: Relevance
“…(4) To enhance model generalization and accuracy, we propose using AR to train the model with adversarial perturbations, improving the model's robustness to minor variations in input data and obtaining a more robust model. (5) Extensive experiments on the TLDITRD dataset confirm that our proposed LAFANet achieves accuracies of 81.7%, 83.3%, and 90.0% for image-to-text retrieval and 80.3%, 93.7%, and 96.3% for text-to-image retrieval, respectively. LAFANet surpasses the baseline models for FNE, advanced VSE∞, and NAAF techniques, boosting R@1 accuracy by 2.4%, 4.6%, and 4.3%, respectively, for image-to-text retrieval, and by 2.3%, 3.0%, and 3.2% for text-to-image retrieval tasks.…”
Section: Discussionmentioning
confidence: 64%
See 1 more Smart Citation
“…(4) To enhance model generalization and accuracy, we propose using AR to train the model with adversarial perturbations, improving the model's robustness to minor variations in input data and obtaining a more robust model. (5) Extensive experiments on the TLDITRD dataset confirm that our proposed LAFANet achieves accuracies of 81.7%, 83.3%, and 90.0% for image-to-text retrieval and 80.3%, 93.7%, and 96.3% for text-to-image retrieval, respectively. LAFANet surpasses the baseline models for FNE, advanced VSE∞, and NAAF techniques, boosting R@1 accuracy by 2.4%, 4.6%, and 4.3%, respectively, for image-to-text retrieval, and by 2.3%, 3.0%, and 3.2% for text-to-image retrieval tasks.…”
Section: Discussionmentioning
confidence: 64%
“…However, manual methods suffer from drawbacks such as high workload, low retrieval efficiency, inconsistent retrieval quality, and reliance on subjective judgment. In the agricultural domain, with the maturation of deep learning technology, complex single-modal management of tomato leaf diseases has been achieved [3][4][5][6][7][8][9][10][11]. Recently, with the diversification of agricultural data types [12], scholars are increasingly inclined towards employing cross-modal information retrieval techniques through various modalities, which provides valuable insights for our research.…”
Section: Introductionmentioning
confidence: 99%
“…First-order and second-order derivative processing are effective methods for resolving overlapping peaks and eliminating background noise [ 24 ]. CWT has the property of refining weak information to highlight localization and can effectively enhance spectrum feature information, and has been applied in noise reduction, de-contextualization, and compression of spectrum data [ 25 , 26 ]. The decomposition scales in CWT were set to 2, , , , , , , , and , which were chosen as the mother wavelet functions for the identification of the dopant content spectra.…”
Section: Methodsmentioning
confidence: 99%
“…The decomposition scales in CWT were set to 2, , , , , , , , and , which were chosen as the mother wavelet functions for the identification of the dopant content spectra. For simplification, the nine decomposition scales were designated as L1–L9, respectively [ 25 , 26 ].…”
Section: Methodsmentioning
confidence: 99%
“…Continuous wavelet transform (CWT) is a mathematical tool used for analyzing non-stationary signals, providing time-frequency representations of these signals by decomposing them into scaled and translated wavelets [19]. CWT convolves an input signal with wavelets at different scales and positions, measuring their similarities to the signal at each scale and time position throughout the convolution process.…”
Section: Introductionmentioning
confidence: 99%