Identification of Traditional Medicinal Plant Leaves Using an effective Deep Learning model and Self-Curated Dataset

Debasish Dutta; Deepjyoti Chetia; Prof Partha Pratim Baruah; Sanjib Kr Kalita; Tanaz Akhter

REVIEW 5 major objections 6 minor 22 references

Identification of Traditional Medicinal Plant Leaves Using an effective Deep Learning model and Self-Curated Dataset

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A compact six-layer CNN identifies medicinal leaves at over 99 percent accuracy across three datasets.

desk verdict The reported accuracies are likely inflated by pre-split data augmentation; the paper's only real asset, the self-curated dataset, is neither released nor properly evaluated. read the letter →

arxiv 2501.09363 v1 pith:SLIMFQMB submitted 2025-01-16 cs.CV

Deepjyoti Chetia , Sanjib Kr Kalita , Prof Partha Pratim Baruah , Debasish Dutta , Tanaz Akhter This is my paper · ORCID

classification cs.CV

keywords medicinalplantidentificationleafclassificationconvolutionalneuralnetworkself-curateddatasetAssamplantsimageaugmentationoptimizercomparisonAyurveda

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a comparatively small convolutional neural network, built from scratch with six convolution layers, can identify medicinal plant leaves as accurately as much heavier transfer-learning models. The authors report test accuracies of 99.55% on 20 classes of the Indian Medicinal Leaves dataset, 98.49% on 10 classes of MED117, and 99.74% on 10 classes of their own newly curated dataset of 50 medicinal plants photographed in Assam. They also contribute the self-curated collection, roughly 42,250 images across 50 species, and compare three optimizers (Adam, RMSprop, SGD with momentum) to show the result is not tied to one training rule. If the claim holds, reliable leaf identification no longer requires an expert botanist at the point of collection, which matters for safe harvesting of Ayurvedic material.

What carries the argument

The load-bearing object is the six-layer CNN itself: six 3x3 convolution layers (32 filters in the first, 64 in the rest) each followed by 2x2 max-pooling, then flatten, a 64-unit dense layer with ReLU, batch normalization, dropout 0.1, and a softmax classifier. It carries the argument by showing that a deliberately simple architecture, trained from scratch rather than fine-tuned, can extract the leaf-texture features needed to separate visually similar medicinal species. The other load-bearing pieces are the three datasets (two public, one newly curated) and the augmentation pipeline (vertical/horizontal flip, zoom to 1.0, 10-degree anticlockwise rotation) that produces four training images from each original.

What would settle it

Retrain the same architecture on the full Indian Medicinal Leaves dataset (80 classes), MED117 (115 classes), and the self-curated collection (50 classes) with the same split, augmentation, and optimizers; if any full-scale accuracy falls substantially below the reported 99.55/98.49/99.74, the central claim as stated for the datasets is refuted. A reader can also request the self-curated data (available upon request) and enumerate the chosen subsets to check whether the selected classes are the easiest.

Watch

Extended reading notes

Core claim

The authors' central claim is that a custom CNN with six 3x3 convolution layers, ReLU activations, max pooling after each convolution, a 64-unit dense layer with batch normalization and 10% dropout, and a softmax output can classify medicinal leaves with high accuracy across datasets of different provenance. Trained from scratch on 256x256 RGB images with random flip, zoom, and 10-degree rotation augmentation and an 80/10/10 split, the model reached 99.55% on a 20-class subset of Indian Medicinal Leaves, 98.49% on a 10-class subset of MED117, and 99.74% on a 10-class subset of the self-curated dataset, with the best results obtained using Adam and RMSprop. On the Indian dataset the model's 99% is comparable to the best previously reported values (98.8% for ECA-VGG19 and 99.71% for MTJNet), which the authors take as evidence that a simple from-scratch network can stand in for pretrained architectures. They also observe that accuracy improves with dataset size, pointing to the value of their curated collection.

Load-bearing premise

The headline numbers come from small, unstated subsets (10 of 50, 20 of 80, and 10 of 115 classes), and the paper assumes those subsets are representative of the full identification problem, so if the omitted classes are harder, the reported accuracies will not generalize to the complete datasets.

Editorial extensions

If this is right

A lightweight from-scratch CNN can reach accuracy comparable to pretrained transfer-learning models, so computational cost need not be a barrier to deploying leaf identification in the field.
The 42,250-image, 50-species self-curated Assam dataset gives the community, if released, a new resource for training and evaluating medicinal-plant classifiers in a region with high ethnobotanical diversity.
Because the model's accuracy improves with dataset size in the authors' experiments, expanding per-class image counts is a direct route to further gains.
On the Indian Medicinal Leaves dataset, the proposed model at 99% is within about one point of the strongest prior result (99.71%), suggesting the simpler architecture is a competitive baseline for future work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: testing on all classes of each dataset, not just subsets, would likely yield lower accuracy because adding visually similar species increases inter-class confusion; the paper's subset choice is unexplained, so the generalization gap is unknown.
Editorial inference: the self-curated dataset was shot under controlled conditions (leaves on white paper, daylight, a high-end smartphone), so real-world accuracy on cluttered or low-light field images is an open question that a mobile-app follow-up would need to answer.
Editorial inference: a direct comparison on the same full datasets, with the same splits, between this six-layer CNN and the cited transfer-learning baselines would settle whether the accuracy advantage is due to architecture, augmentation, or the particular class subsets used.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The thing to know about this paper: the headline accuracies (99.55%, 98.49%, 99.74%) are probably a leakage artifact. Section 3.1 describes generating four augmented images per source leaf, and says augmentation is applied to 'the dataset'. The 80/10/10 train/test/validation split is described in Section 3.2, after augmentation is already done. The text never says augmentation was restricted to the training split. So augmented copies of the same original leaf can easily end up in both train and test. That makes the test set measure near-duplicate recognition, not generalization.

What's genuinely new here is the self-curated dataset: 42,250 images of 50 Assam medicinal plants, captured with an iPhone 14 Pro. That could be useful to the community. But only 10 of the 50 classes are used for evaluation, and the dataset is 'available upon request' rather than deposited. So the contribution is largely a promise.

What the paper does well is a straightforward comparison of three optimizers (Adam, RMSprop, SGD with momentum) on the same CNN, with precision/recall/F1 reported. The related work section is adequate. The architecture is a stock six-conv-layer CNN; no new layer, loss, or training scheme is proposed.

Beyond the leakage ambiguity, the subset selection is unjustified: 10 of 50, 20 of 80, 10 of 115 classes, with no enumeration or criterion. The abstract reports the accuracies without mentioning subsets, which overstates coverage. There is a single run, no error bars, and no code release. Table 3 lists 'Proposed 99%' while Table 2 says 99.55% on the Indian dataset, a small inconsistency.

If the augmentation was actually done after the split, the authors should say so explicitly and re-run with a clarified protocol. If not, the accuracies are invalid. Either way, the paper as written can't be trusted.

This is a paper for a workshop or a methods note at best. I would not send it to peer review in its current form; the leakage risk is a load-bearing flaw, and the dataset is not accessible to verify anything. I would, however, encourage the authors to release the dataset and fix the protocol, because a clean evaluation on the full self-curated data would be worth reading.

Referee Report

5 major / 6 minor

Summary. The paper proposes a custom convolutional neural network with six convolution layers for the identification of traditional medicinal plant leaves, and evaluates it on three datasets: the authors' self-curated dataset, the Indian Medicinal Leaves Image Dataset, and the MED117 dataset. The model is reported to achieve test accuracies of 99.55%, 98.49%, and 99.74% respectively (Table 2), with the self-curated dataset containing 42,250 images of 50 plant species. The authors compare several optimizers (Adam, RMSprop, SGD with momentum) and claim the proposed model is effective for medicinal plant identification. The evaluation, however, is performed on subsets of each dataset, and the experimental protocol as described leaves open the possibility of data leakage from pre-split augmentation.

Significance. If the reported results were valid, the paper would provide a simple and effective CNN baseline for medicinal plant leaf classification, along with a potentially useful self-curated dataset for a region (Assam, India) that is underrepresented in existing benchmarks. The practical motivation is clear, and the reported accuracy numbers are competitive with prior work. However, the current evidence for these claims is undermined by experimental-protocol ambiguities and reporting gaps: the augmentation pipeline appears to be applied before the train/test split, the evaluation uses only subsets of the full datasets without justification, results are single-run with no error bars, and the abstract presents subset accuracies as though they apply to the full datasets. These issues must be resolved before the central claims can be accepted. The paper does not release code or the self-curated dataset publicly, which further limits reproducibility.

major comments (5)

[Section 3.1, Section 3.2] The data augmentation pipeline is described in Section 3.1 ('Data augmentation is applied to the dataset to avoid overfitting... four images in total were generated from one pre-processed plant leaf image'), and only in Section 3.2 is the dataset split into 80% train, 10% test, and 10% validation. The text never states that augmentation is restricted to the training set. If augmentation is performed before the split, augmented copies of the same original image can appear in both the training and test sets, so the accuracies in Table 2 would reflect near-duplicate recognition rather than generalization to unseen images. This is a load-bearing issue because the central claim of high identification accuracy depends entirely on the validity of the held-out test set. The authors must clarify the order of operations and, if augmentation preceded the split, rerun all experiments with augmentation applied only to the training partition.
[Section 3.1, Abstract, Table 2] The model is evaluated on subsets of each dataset: 10 of 50 self-curated classes, 20 of 80 Indian Medicinal Leaves classes, and 10 of 115 MED117 classes. No selection criterion is provided, and the chosen classes are not enumerated. The abstract reports the resulting accuracies (99.5%, 98.4%, 99.7%) without any qualification, implying that the full datasets were tested. This overstates the coverage of the results. The authors should either evaluate on the full datasets or explicitly state in the abstract and conclusion that the results are for subsets, and they must justify that the chosen subsets are representative of the full datasets and of real-world identification difficulty.
[Section 4, Table 2] The reported accuracies are from a single experimental run with no standard deviations, confidence intervals, or repeated-seed information. The model and optimizer are effectively selected using validation accuracy, and the test accuracy is reported without accounting for this selection. Single-run results of this kind are not sufficient to establish that the proposed model reliably achieves the claimed accuracy. The authors should report the mean and standard deviation over multiple runs (or at least provide the results from several independent runs), and they should describe the model-selection procedure clearly.
[Section 4, Section 6] The text is inconsistent about which optimizer produced the reported results. Section 4 states 'RMSprop showed the best results' and later 'Right after Adam, RMSprop and SGD (Momentum) performed best,' while the conclusion says 'Adam and RMSprop have performed best in all three datasets.' Table 2 lists only one accuracy per dataset, without indicating the optimizer or epoch that produced it. The authors must specify, for each dataset, the optimizer and epoch count corresponding to the numbers in Table 2, and reconcile the claims about optimizer performance.
[Table 3] The comparison in Table 3 is misleading as presented. The proposed method is compared with prior models on the Indian Medicinal Leaves dataset, but the proposed method was evaluated on a 20-class subset, whereas the prior works (e.g., MTJNet, ECA-VGG19) report results on the full 80-class dataset. This is not a like-for-like comparison. The table should clearly indicate the number of classes used for each method, or the comparison should be restricted to methods evaluated under the same protocol.

minor comments (6)

[Throughout] There are numerous typographical and grammatical errors, including 'trainning' (Section 4), 'achived' (Section 4), 'eopch' (Section 4), 'image suze' (Section 4), 'Ensamble' (Table 3), 'mages' (Figure 2 caption), and 'Indian Medicinal Leaves image' (Section 3.1). A careful proofreading pass is needed.
[Section 3.3] The statement 'we have not set any learning rate, and the default value of 0.001 is considered for all the optimizers' is not accurate for standard frameworks; for example, Keras's SGD default learning rate is 0.01 while Adam and RMSprop default to 0.001. The authors should specify the deep learning framework and the exact hyperparameter values used for each optimizer.
[Figure 5] The subfigures showing training and validation accuracy versus epoch do not indicate which optimizer they correspond to. Since the paper compares Adam, RMSprop, and SGD with momentum, each curve should be labeled with the optimizer or the figure should be split accordingly.
[Section 1.3, Section 3.1] The contribution states that the self-curated dataset contains 42,250 images of 50 medicinal plants, but Section 3.1 reports that the class sizes range from 110 to 1800 images. The total of 42,250 is inconsistent with the stated per-class range (50 classes at an average of 845 images each is possible, but the range 110-1800 suggests the total would be between 5,500 and 90,000; please clarify the exact distribution and confirm the total count).
[Table 1] The related-work table contains inconsistencies with the text: for example, reference [16] is described in the text as using 1500 images of 40 plants, but the table says 500 images of 40 plants; reference [19] is called both 'SDAMPI' and 'SDAMPI' in different places. These discrepancies should be corrected.
[Section 7.1] The self-curated dataset is said to be 'available upon request,' which is not a practical form of data availability for reproducibility. The authors should upload the dataset to a public repository, as they have done for the other two datasets.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the reported accuracies are empirical held-out test results, not derived quantities reducible to the inputs.

full rationale

The paper's central claim is an experimental measurement: a custom CNN is trained from scratch on three datasets, and accuracy, precision, recall, and F1 are reported on a held-out test split (Section 4, Table 2). This is not a derivation in which a predicted quantity equals, by construction, a fitted parameter or a self-cited theorem. The model architecture, optimizers, epochs, and batch size are chosen through standard training and validation practice, and the external datasets (Indian Medicinal Leaves Image Dataset [10] and MED117 [9]) serve as independent benchmarks. The main methodological risks, such as the ambiguous ordering of data augmentation relative to the split and the unstated criteria for selecting class subsets, concern the validity and generalizability of the measurements, not circularity: the reported test accuracy is still computed on images held out from the training process rather than being defined to match the training objective. No self-citation is load-bearing, and no 'prediction' reduces to an input by construction. Therefore the appropriate circularity score is 0.

Assumptions & free parameters 8 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the quality and representativeness of the three datasets and subsets, plus standard CNN hyperparameters chosen by hand. The paper provides no independent evidence for the representativeness of the subsets, and the self-curated dataset is not public.

free parameters (8)

image_size = 256x256
All images resized to 256x256 before training (Section 3.2).
batch_size = 32
Chosen as best among several examined (Section 3.3).
learning_rate = 0.001
Default value used for all optimizers (Section 3.3).
dropout_rate = 0.1
Set to regularize after the dense layer (Section 3.3).
num_conv_layers = 6
Architecture choice; first layer 32 filters, subsequent 64 (Section 3.3).
dense_units = 64
Fully connected layer size (Section 3.3).
class_subset_choices = 10/20/10 classes
Subsets taken from the three datasets without justification (Section 3.1).
epochs = 8 to 65 depending on dataset/optimizer
Chosen by increasing until validation accuracy stops improving (Section 4).

assumptions (4)

domain assumption The publicly available datasets are correctly labeled.
The paper relies on the labels provided by the Mendeley dataset creators without independent verification (Section 3.1).
ad hoc to paper The chosen subsets (10 of 50, 20 of 80, 10 of 115 classes) are representative of the full datasets and real-world difficulty.
No criterion for subset selection is given; selection could favor easily separable classes (Section 3.1).
domain assumption The train/validation/test split prevents leakage of near-duplicate images of the same leaf or plant.
The paper does not describe deduplication or whether multiple images of the same physical leaf appear in different splits (Section 3.2).
domain assumption Controlled capture conditions (white paper background, iPhone 14 Pro) are representative of field use.
The self-curated dataset is captured under fixed conditions, yet the stated use case is identification by laypeople in the field (Section 3.1).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Identification of Traditional Medicinal Plant Leaves Using an effective Deep Learning model and Self-Curated Dataset." pith.science (2026). https://pith.science/paper/SLIMFQMB

@misc{pith2026250109363,
  author       = {Pith},
  title        = {Pith review of: Identification of Traditional Medicinal Plant Leaves Using an effective Deep Learning model and Self-Curated Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SLIMFQMB}},
  note         = {Machine review of arXiv:2501.09363}
}

read the original abstract

Medicinal plants have been a key component in producing traditional and modern medicines, especially in the field of Ayurveda, an ancient Indian medical system. Producing these medicines and collecting and extracting the right plant is a crucial step due to the visually similar nature of some plants. The extraction of these plants from nonmedicinal plants requires human expert intervention. To solve the issue of accurate plant identification and reduce the need for a human expert in the collection process; employing computer vision methods will be efficient and beneficial. In this paper, we have proposed a model that solves such issues. The proposed model is a custom convolutional neural network (CNN) architecture with 6 convolution layers, max-pooling layers, and dense layers. The model was tested on three different datasets named Indian Medicinal Leaves Image Dataset,MED117 Medicinal Plant Leaf Dataset, and the self-curated dataset by the authors. The proposed model achieved respective accuracies of 99.5%, 98.4%, and 99.7% using various optimizers including Adam, RMSprop, and SGD with momentum.

Figures

Figures reproduced from arXiv: 2501.09363 by the authors.

**Figure 1.** Proposed methodology 3.1 Dataset preparation Data collection In this study, 3 datasets were used to evaluate the model’s performance. Two of them were collected from publicly available domains and one was curated by the authors. The first dataset that was selected for evaluation is curated by the authors. Images were collected from the Botanical Garden, Gauhati University, and Kaziranga Orchid Nursery in Bokakhat, … view at source ↗

**Figure 2.** mages of the self-curated dataset camera angles. Plant leaves were captured in daylight with illuminance range 32,000–100,000 lux. The dataset also has few blurred noisy images. These variations in image quality is common in real world scenarios. The dataset contains 50 medicinal plant types and no of images in each class ranges from 110 to 1800. During dataset preparation, the image dimension was reduced using a P… view at source ↗

**Figure 3.** Image data augmentation process Data augmentation Data augmentation is applied to the dataset to avoid overfitting. For generating augmented images, augmentation techniques, i.e. random flip, random zoom, and random rotation, were applied. Images are flipped vertically and horizontally. All the images are zoomed in at 1.0. Medicinal plant leaf images are rotated at 10◦ in an anti-clockwise manner. In this way, fou… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Proposed CNN architecture Parameter Setting for the proposed model To update the different parameters, the training process of the proposed model has been evaluated iteratively. Some important parameters are batch size, regularization parameter, number of epochs, and …

**Figure 5.** Figure 5: Train and Validation accuracy vs. epoch of the model on three datasets [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 17 canonical work pages

[1]

Metabolites9 (2019)

Anand, U., Jacobo-Herrera, N., Altemimi, A., Lakhssassi, N.: A Comprehensive Review on Medicinal Plants as Antimicrobial Therapeutics: Potential Avenues of Biocompatible Drug Discovery. Metabolites9 (2019). https://doi.org/10.3390/ metabo9110258

work page 2019
[2]

Agronomy 12 (2022)

Azadnia, R., Al-Amidi, M., Mohammadi, H., Cifci, M., Daryab, A., Cavallo, E.: An AI Based Approach for Medicinal Plant Identification Using Deep CNN Based on Global Average Pooling. Agronomy 12 (2022). https://doi.org/10.3390/ agronomy12112723

work page 2022
[3]

Information (Switzerland) 14 (2023)

Hajam, M., Arif, T., Khanday, A., Neshat, M.: An Effective Ensemble Convolu- tional Learning Model with Fine-Tuning for Medicinal Plant Leaf Identification. Information (Switzerland) 14 (2023). https://doi.org/10.3390/info14110618

work page doi:10.3390/info14110618 2023
[4]

International Book Distributors, Dehradun, Uttarak- hand, India (2009)

Handique,P.J.(ed.): Medicinalplantsof NorthEast India:status, diversity, conser- vation, cultivation & trade. International Book Distributors, Dehradun, Uttarak- hand, India (2009)

work page 2009
[5]

Kingma, D.P., Ba, J.: Adam: A Method for Stochastic Optimization (2017)

2017
[6]

https://doi.org/10.2139/ssrn.4804339

K.R, B., Rani, N.S., Devadas, R., Hiremani, D.V.: Enhancing Medicinal Plant Leaves Species Classification Through Ensemble Learning with Transfer Learning Integration (2024). https://doi.org/10.2139/ssrn.4804339

work page doi:10.2139/ssrn.4804339 2024
[7]

IEEE Transactions on Neural Net- works and Learning Systems33 (2022)

Li, Z., Liu, F., Yang, W., Peng, S., Zhou, J.: A Survey of Convolutional Neural Net- works: Analysis, Applications, and Prospects. IEEE Transactions on Neural Net- works and Learning Systems33 (2022). https://doi.org/10.1109/TNNLS.2021. 3084827

doi:10.1109/tnnls.2021 2022
[8]

12 11 (2023)

Miao, J., Huang, Y., Wang, Z., Wu, Z., Lv, J.: Image recognition of traditional Chinese medicine based on deep learning. 12 11 (2023). https://doi.org/10. 3389/fbioe.2023.1199803

arXiv 2023

Show all 22 references

[9]

https://doi.org/10.17632/DTVBWRHZNZ.4

Parismita Sarma: MED117_medicinal Plant Leaf Dataset & Name Table (2023). https://doi.org/10.17632/DTVBWRHZNZ.4

2023 doi
[10]

org/10.17632/748F8JKPHB.3

Pushpa Gangadkar: Indian Medicinal Leaves Image Datasets (2023).https://doi. org/10.17632/748F8JKPHB.3

2023 doi
[11]

https://doi.org/10.1109/ GCITC60406.2023.10426044

Rohith, N., Bhuvaneswari, R.: Medicinal plant classification using attention mechanism driven VGG19 and Resnet101 (2023). https://doi.org/10.1109/ GCITC60406.2023.10426044

2023
[12]

IEEE Access 9 (2021)

Roopashree, S., Anitha, J.: DeepHerb: A Vision Based System for Medicinal Plants Using Xception Features. IEEE Access 9 (2021). https://doi.org/10. 1109/ACCESS.2021.3116207

2021
[13]

Multimedia Tools and Applications (2023).https: //doi.org/10.1007/s11042-023-17311-8

Sachar, S., Kumar, A.: A novel transfer learning-based approach for plant species prediction using leaf images. Multimedia Tools and Applications (2023).https: //doi.org/10.1007/s11042-023-17311-8

2023 doi
[14]

In: Saponara, S., De Gloria, A

Saponara, S., Elhanashi, A.: Impact of Image Resizing on Deep Learning Detec- tors for Training Time and Model Performance. In: Saponara, S., De Gloria, A. (eds.) Applications in Electronics Pervading Industry, Environment and Society, vol. 866. Springer International Publishi...

2022
[15]

Schmidt, R.M., Schneider, F., Hennig, P.: Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers (2021)

2021
[16]

Multimedia Tools and Applications (2023)

Sharma, M., Kumar, N., Sharma, S., Kumar, S., Singh, S., Mehandia, S.: Medicinal plants recognition using heterogeneous leaf features: an intelligent ap- proach. Multimedia Tools and Applications (2023). https://doi.org/10.1007/ s11042-023-17639-1 Identification of Traditional...

2023
[17]

Knowledge-Based Systems (2024).https: //doi.org/https://doi.org/10.1016/j.knosys.2024.112147

Sharma, S., Vardhan, M.: Mtjnet: Multi-task joint learning network for advancing medicinal plant and leaf classification. Knowledge-Based Systems (2024).https: //doi.org/https://doi.org/10.1016/j.knosys.2024.112147

2024
[18]

Multimedia Tools and Applications82(2023)

Thanikkal, J., Dubey, A., Thomas, M.: Deep - Morpho Algorithm (DMA) for medicinal leaves features extraction. Multimedia Tools and Applications82(2023). https://doi.org/10.1007/s11042-023-14567-y

2023 doi
[19]

Wireless Personal Communications 131 (2023)

Thanikkal, J., Dubey, A., Thomas, M.: An Efficient Mobile Application for Iden- tification of Immunity Boosting Medicinal Plants using Shape Descriptor Al- gorithm. Wireless Personal Communications 131 (2023). https://doi.org/10. 1007/s11277-023-10476-3

2023
[20]

MATHEMATICS11(2023)

Uddin, A., Chen, Y., Borkatullah, B., Khatun, M., Ferdous, J., Mahmud, P., Yang, J., Ku, C., Por, L.: Deep-Learning-Based Classification of Bangladeshi Medicinal Plants Using Neural Ensemble Models. MATHEMATICS11(2023). https://doi. org/10.3390/math11163504

2023 doi
[21]

Journal of Intelligent and Fuzzy Systems42(2022)

Uma, K., Sathya Bama, B., Sabarinathan, D., Mansoor Roomi, S.: Identification and retrieval of medicinal plants of Southern India using EfficientB4Net. Journal of Intelligent and Fuzzy Systems42(2022). https://doi.org/10.3233/JIFS-211426

2022 doi
[22]

Woo, S., Park, J., Lee, J.Y., Kweon, I.S.: CBAM: Convolutional Block Attention Module (2018)

2018

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

Metabolites9 (2019)

Anand, U., Jacobo-Herrera, N., Altemimi, A., Lakhssassi, N.: A Comprehensive Review on Medicinal Plants as Antimicrobial Therapeutics: Potential Avenues of Biocompatible Drug Discovery. Metabolites9 (2019). https://doi.org/10.3390/ metabo9110258

work page 2019

[2] [2]

Agronomy 12 (2022)

Azadnia, R., Al-Amidi, M., Mohammadi, H., Cifci, M., Daryab, A., Cavallo, E.: An AI Based Approach for Medicinal Plant Identification Using Deep CNN Based on Global Average Pooling. Agronomy 12 (2022). https://doi.org/10.3390/ agronomy12112723

work page 2022

[3] [3]

Information (Switzerland) 14 (2023)

Hajam, M., Arif, T., Khanday, A., Neshat, M.: An Effective Ensemble Convolu- tional Learning Model with Fine-Tuning for Medicinal Plant Leaf Identification. Information (Switzerland) 14 (2023). https://doi.org/10.3390/info14110618

work page doi:10.3390/info14110618 2023

[4] [4]

International Book Distributors, Dehradun, Uttarak- hand, India (2009)

Handique,P.J.(ed.): Medicinalplantsof NorthEast India:status, diversity, conser- vation, cultivation & trade. International Book Distributors, Dehradun, Uttarak- hand, India (2009)

work page 2009

[5] [5]

Kingma, D.P., Ba, J.: Adam: A Method for Stochastic Optimization (2017)

2017

[6] [6]

https://doi.org/10.2139/ssrn.4804339

K.R, B., Rani, N.S., Devadas, R., Hiremani, D.V.: Enhancing Medicinal Plant Leaves Species Classification Through Ensemble Learning with Transfer Learning Integration (2024). https://doi.org/10.2139/ssrn.4804339

work page doi:10.2139/ssrn.4804339 2024

[7] [7]

IEEE Transactions on Neural Net- works and Learning Systems33 (2022)

Li, Z., Liu, F., Yang, W., Peng, S., Zhou, J.: A Survey of Convolutional Neural Net- works: Analysis, Applications, and Prospects. IEEE Transactions on Neural Net- works and Learning Systems33 (2022). https://doi.org/10.1109/TNNLS.2021. 3084827

doi:10.1109/tnnls.2021 2022

[8] [8]

12 11 (2023)

Miao, J., Huang, Y., Wang, Z., Wu, Z., Lv, J.: Image recognition of traditional Chinese medicine based on deep learning. 12 11 (2023). https://doi.org/10. 3389/fbioe.2023.1199803

arXiv 2023