A Hybrid ConvNeXt-EfficientNet AI Solution for Precise Falcon Disease Detection

Alavikunhu Panthakkan; Fatma Taher; Hussain Al-Ahmad; S M Anzar; Zubair Medammal

REVIEW 5 major objections 6 minor 12 references

A Hybrid ConvNeXt-EfficientNet AI Solution for Precise Falcon Disease Detection

T0 review · 5 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that concatenating ConvNeXt and EfficientNet into one classifier labels falcon endoscopy images as normal, liver disease, or aspergillosis with 98% reported test accuracy, beating the two base models by 4 and 11 points.

desk verdict New dataset and domain, but the 98% claim rests on a split protocol the paper never actually describes. read the letter →

arxiv 2506.14816 v1 pith:EVEP4VNH submitted 2025-06-08 cs.CV

Alavikunhu Panthakkan , Zubair Medammal , S M Anzar , Fatma Taher , Hussain Al-Ahmad This is my paper

classification cs.CV

keywords falcondiseasedetectionConvNeXtEfficientNetconcatenatedmodelendoscopyimageclassificationaspergillosislivertransferlearning

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 tries to establish that a hybrid deep-learning model, built by concatenating the ConvNeXt and EfficientNet image classifiers, can sort falcon endoscopy images into three categories—normal, liver disease, and aspergillosis—more accurately than either architecture on its own. On a dataset of 610 labeled images from the Sharjah Falcon Clinic, the authors report an average test accuracy of 98% for the hybrid, against 87% for ConvNeXt and 94% for EfficientNet. The significance of the claim, if it holds, is that automated image triage could give falconers and veterinarians a fast, consistent second opinion for two serious avian conditions. The paper also frames the result as a step toward AI-assisted avian healthcare beyond falcons.

What carries the argument

The load-bearing object is the concatenation operation itself: a model architecture whose final feature representations from ConvNeXt and from EfficientNet are merged and passed to a classifier with three output classes. The paper sets the input at 128×128 pixels and trains the joined model over 50 epochs on 80% of the data, using transfer learning to initialize the two branches. ConvNeXt supplies fine-grained visual pattern extraction; EfficientNet supplies efficient scaling and robust feature reuse. The mechanism works by letting each branch's strengths compensate for the other's weaknesses, which is why the authors expect the combined score to beat either branch alone.

What would settle it

Rebuild the model on the same 610 images using three disjoint splits—train, validation for tuning, and a test set seen only once at the end. If the test accuracy on that untouched split falls well below 98%, or no longer clearly beats ConvNeXt and EfficientNet alone, the paper's central claim fails. A cheaper check is to inspect whether any of the 126 test images used for the reported confusion matrices also appear in the 20% set described as 'internal validation.'

Watch

Extended reading notes

Core claim

The central claim is that a "concatenated AI model"—the feature outputs of ConvNeXt and EfficientNet joined into one classifier, fine-tuned on endoscopy images—classifies Normal, Liver Disease, and Aspergillosis with a reported 98% average testing accuracy, 0.97 precision, 0.98 recall, and 0.98 F1-score. Per-class results in the paper's Table I show the hybrid reaching 0.96, 0.99, and 0.98 accuracy for Normal, Liver, and Aspergillosis respectively, with the weakest per-class cell being Liver recall at 0.89 for EfficientNet and 0.83 for ConvNeXt. The authors argue this hybrid outperforms traditional diagnostic methods and the two base models on the same test images. In the authors' words, the concatenated model "achieves testing accuracies of 98%, surpassing other deep learning models considered in this study."

Load-bearing premise

The reported 98% accuracy depends on the test images being truly unseen: if the same 20% of images used to tune hyperparameters was also used to compute the final metrics, the score is not an honest estimate.

Editorial extensions

If this is right

If the 98% accuracy holds, the model can automatically label the great majority of falcon endoscopy images as normal, liver disease, or aspergillosis without a specialist viewing each frame.
The hybrid's gains over ConvNeXt (87%) and EfficientNet (94%) imply that combining the two feature extractors extracts information neither branch captures alone on this dataset.
The paper's transfer-learning approach means the same two-branch architecture can be fine-tuned on new disease classes with far fewer images than training from scratch would require.
A veterinary clinic could use the model as a screening step, flagging suspected aspergillosis or liver disease for rapid confirmatory tests.

Reading between the lines

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

I would infer that this concatenation recipe is a general template for veterinary image triage, not a falcon-specific finding, but the paper only demonstrates it on one dataset from one clinic.
I would not expect the 98% figure to transfer unchanged to images from other clinics, cameras, or lighting conditions, because the paper reports no cross-clinic evaluation.
The clearest testable extension is to run the identical three-way split with a validation set that is never shown during hyperparameter search; if accuracy stays near 98%, the claim is solid, and if it drops, the published number was inflated by tuning on the test set.

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

Quick take: the paper is a straightforward reuse of the authors' prior concatenated ConvNeXt-EfficientNet architecture, applied to a new, small falcon endoscopy dataset. The new dataset and the three-class problem (Normal, Liver, Aspergillosis) are genuinely new and could be useful to avian veterinarians, but the reported 98% test accuracy cannot be trusted as an unbiased number because the paper never cleanly separates validation and test.

What it does well: it assembles a labeled dataset of 610 endoscopy images from a falcon clinic, and it compares the hybrid against the two base architectures. The performance table is coherent in showing the hybrid above the baselines, and the authors are transparent about their prior use of the same concatenation in [2]. That kind of direct extension is fine, provided the evaluation is sound.

The soft spots are real and load-bearing. The pseudocode in Section III-A defines only an 80/20 train/test split. Section IV then says the remaining 20% was used for 'internal validation' and that hyperparameter tuning was done on the validation set, while also reporting metrics on a 'test set' that is never separately described. The numbers also do not reconcile: 20% of 610 images is 122, but the confusion matrices are said to cover 126 test images. On top of that, the 'Accuracy' column in Table I lists per-class values without any definition; that is unusual and makes the reported average uninterpretable without the raw confusion matrices. No code or data are provided, and there are no error bars or repeated-seed statistics. None of this proves the method is bad, but the evaluation as written does not support the conclusion's claim of 'testing accuracies of 98%' as a reliable estimate.

Who is this for? Avian veterinarians and applied-ML folks working on small medical imaging datasets. It is a niche contribution, not a methodological advance. A serious referee could help the authors tighten the split protocol, define the metrics, and release code/data. With that, the paper would be a modest but legitimate addition to the literature. As is, it should go to peer review rather than desk reject, because the dataset is new and the claim is potentially valuable but unverified.

Recommendation: send to peer review, with instructions to the authors to clarify the split, provide the actual test set composition, define per-class accuracy, and share code and data.

Referee Report

5 major / 6 minor

Summary. The paper proposes a hybrid deep-learning model that concatenates features from ConvNeXt and EfficientNet to classify falcon endoscopy images into three classes: Normal, Liver Disease, and Aspergillosis. The authors report an average test accuracy of 98% for the hybrid model, compared with 87% for ConvNeXt and 94% for EfficientNet, using a dataset of 610 images from the Sharjah Falcon Clinic. The manuscript includes training curves, confusion matrices, and an ROC curve, and concludes that the concatenated model surpasses both individual architectures and traditional diagnostic methods.

Significance. If the reported accuracy were obtained on a genuinely held-out test set, this would be a modest but useful application of transfer learning to a niche veterinary imaging problem. The comparison against two strong baselines is a reasonable starting point, and the paper gives credit for reporting confusion matrices and training curves. However, the claimed 4-point improvement over EfficientNet rests on a small, single-split, closed dataset with no statistical analysis, and the evaluation protocol is ambiguous. The architectural contribution is also incremental, reusing the concatenation scheme from the authors' prior brain-tumor classification work [2]. The current significance is therefore unverified and largely empirical.

major comments (5)

[Section III-A and Section IV] The evaluation protocol is ambiguous: the pseudocode in Section III-A describes only a training (80%) and testing (20%) split, while Section IV states that the remaining 20% is used for internal validation and that hyperparameter fine-tuning was executed using the validation set. No separate held-out test set is ever described. If the same 20% of images was used both to tune hyperparameters and to compute the test metrics in Table I, the 98% figure is not an unbiased estimate of generalization, and the comparison against ConvNeXt and EfficientNet is not a fair test.
[Section IV-B] There is an internal inconsistency in the number of test images: 20% of the stated 610 images is 122, but Section IV-B says the confusion matrices are based on 126 test images. The paper does not explain where the extra four images come from or whether the split was not exactly 80/20. This discrepancy undermines the split accounting and makes the reported metrics unverifiable.
[Table I and Section IV-A] The Accuracy column in Table I lists per-class accuracy values and an average, but the manuscript never defines per-class accuracy or states whether the average is a macro average. Equation (1) defines only overall accuracy. Without a precise definition and without the underlying per-class counts, the reported average of 0.98 cannot be checked from the confusion matrices.
[Abstract and Conclusion] The claim that the hybrid model outperforms traditional diagnostic methods is unsupported by the experiments: the study compares only ConvNeXt and EfficientNet and includes no traditional diagnostic modalities such as veterinarian interpretation, radiography, or laboratory tests. This overclaim should be removed or supported by a direct comparative experiment.
[Section IV] No error bars, confidence intervals, or repeated-run statistics are reported for any of the models. With a single 80/20 split and only 610 images, the difference between 0.98 and 0.94 may be within run-to-run variability, so the claimed superiority of the proposed model over EfficientNet is not statistically established.

minor comments (6)

[Section I] There are typographical issues with quotation marks in the abstract and introduction, such as "Normal'", "Liver Disease'", and "'Aspergillosis'", where opening quotes are missing.
[Section IV-B] The text says "Figure 4 and 5" and "Figure 6 show"; the first should be "Figures 4 and 5" and the second should be "Figure 6 shows".
[Equation (4)] The word "Specificity" is misspelled as "Specif icity" in Equation (4).
[Section IV] The paper states that each of the three classes contains 203 images, which sums to 609, but the total dataset is stated as 610 images; the remaining image is not accounted for.
[Fig. 2 and Section III-A] The block diagram is not accompanied by a layer-wise description of the concatenation, such as how the ConvNeXt and EfficientNet features are fused and which dense layers follow the concatenation; specifying this would improve reproducibility.
[Reference [2]] Reference [2] is the authors' own prior paper on a similar concatenated architecture for brain-tumor classification; the manuscript should state explicitly what is new in the present work beyond applying that architecture to a different dataset.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the reported accuracy is an empirical measurement on falcon images, not a quantity forced by the paper's equations or by the self-cited hybrid-architecture paper.

full rationale

The paper makes no first-principles derivation: it trains a concatenation of pretrained ConvNeXt and EfficientNet features on 610 falcon endoscopy images and reports measured accuracy, precision, recall, and F1. The proposed model's superiority is an empirical comparison against two baselines on the same task, not a consequence of the definitions. The only self-citation is [2], the authors' earlier brain-tumor classification paper, which motivates reusing the concatenation ansatz; it is not invoked as a uniqueness theorem or as proof that the falcon result must hold, and the current falcon dataset is independent of that prior work. The 80/20 split language (pseudocode says training/testing; Section IV says training/internal validation with hyperparameter tuning on the validation set) and the mismatch between 122 and 126 test images raise a possible data-leakage or reproducibility concern, but no quoted equation renders the test accuracy equal to the validation accuracy by construction. Under the stated rules, evaluation-bias concerns are correctness risks, not circularity; therefore no circular step is identified.

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

The empirical claim depends on correctly labeled images, transferable pretrained features, and a representative evaluation split. The dataset is small and closed; the paper provides no independent label verification and no code, so these assumptions cannot be checked. No new physical entities or parameters beyond standard ML hyperparameters are introduced.

free parameters (4)

Input image resolution = 128x128
Chosen in Section III-A pseudocode; affects model capacity and the amount of detail visible to the network.
Batch size = 5
Set in Section IV; influences gradient noise and convergence.
Training epochs = 50
Set in Section IV; no early stopping or learning-rate schedule is reported.
Train/test split ratio = 80/20
Pseudo code and Section IV; if the same 20% is used for hyperparameter tuning, reported metrics may be optimistic.

assumptions (4)

domain assumption Endoscopy images are correctly labeled as Normal, Liver, or Aspergillosis.
The paper does not describe a labeling protocol, inter-rater agreement, or confirmatory tests; Section IV simply states 203 images per class from the Sharjah Falcon Clinic.
domain assumption Pretrained ImageNet features are transferable to falcon endoscopy images.
The method relies on transfer learning (Section III) but presents no analysis that ImageNet features are relevant to endoscopic avian images.
domain assumption The 80/20 random split produces an independent test set.
Pseudocode in Section III-A says 80% training and 20% testing; Section IV calls the 20% 'internal validation' and tunes hyperparameters on it. The independence of the test set from model selection is not established.
standard math Standard classification metrics (Equations 1-5) are appropriate for this dataset.
The paper uses standard accuracy, precision, recall, specificity, and F1 definitions in Section IV-A.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Hybrid ConvNeXt-EfficientNet AI Solution for Precise Falcon Disease Detection." pith.science (2026). https://pith.science/paper/EVEP4VNH

@misc{pith2026250614816,
  author       = {Pith},
  title        = {Pith review of: A Hybrid ConvNeXt-EfficientNet AI Solution for Precise Falcon Disease Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EVEP4VNH}},
  note         = {Machine review of arXiv:2506.14816}
}

read the original abstract

Falconry, a revered tradition involving the training and hunting with falcons, requires meticulous health surveillance to ensure the health and safety of these prized birds, particularly in hunting scenarios. This paper presents an innovative method employing a hybrid of ConvNeXt and EfficientNet AI models for the classification of falcon diseases. The study focuses on accurately identifying three conditions: Normal, Liver Disease and 'Aspergillosis'. A substantial dataset was utilized for training and validating the model, with an emphasis on key performance metrics such as accuracy, precision, recall, and F1-score. Extensive testing and analysis have shown that our concatenated AI model outperforms traditional diagnostic methods and individual model architectures. The successful implementation of this hybrid AI model marks a significant step forward in precise falcon disease detection and paves the way for future developments in AI-powered avian healthcare solutions.

Figures

Figures reproduced from arXiv: 2506.14816 by the authors.

**Figure 1.** UAE Falcon AI model significantly surpasses traditional diagnostic methods and individual architectures in terms of accuracy and reliability. The successful implementation of this hybrid AI model in falcon disease detection marks a significant advancement in veterinary science. It not only enhances the precision in diagnosing falcon ailments but also lays the groundwork for future innovations in AI-powered solution… view at source ↗

**Figure 2.** Block Diagram of the Proposed Concatenated ML Model [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Training Performance of the Proposed Model [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Confusion Matrix of the Base Models [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 6.** Figure 6: ROC-AUC Plot of the Proposed Concatenated Model [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

12 extracted references · 11 canonical work pages

[2]

Panthakkan, S

A. Panthakkan, S. Anzar, and W. Mansoor, ``Unleashing the power of efficientnet-convnext concatenation for brain tumor classification,'' in 2023 15th Biomedical Engineering International Conference (BMEiCON). 1em plus 0.5em minus 0.4em IEEE, 2023, pp. 1--5

work page 2023
[1]

Fischer, L

D. Fischer, L. Van Waeyenberghe, C. Cray, M. Gross, E. Usleber, F. Pasmans, A. Martel, and M. Lierz, ``Comparison of diagnostic tools for the detection of aspergillosis in blood samples of experimentally infected falcons,'' Avian Diseases, vol. 58, no. 4, pp. 587--598, 2014

work page 2014
[3]

Muller, T

M. Muller, T. Mannil, and A. George, ``Study on the most common bacterial infections in falcons in the united arab emirates,'' in Proceedings of the 27th Annual AAV Conference, San Antonio, TX, USA, 2006, pp. 6--10

work page 2006
[4]

M. A. A. Hemeiri, A. A. de la Torre, K. Mohteshamuddin, B. A. Degafa, and G. Ameni, ``Retrospective study on the health problems of falcons in al ain, united arab emirates,'' bioRxiv, pp. 2021--02, 2021

work page 2021
[5]

A. F. Alrefaei, ``Molecular detection and genetic characterization of trichomonas gallinae in falcons in saudi arabia,'' Plos one, vol. 15, no. 10, p. e0241411, 2020

work page 2020
[6]

Tarello, ``Etiologic agents and diseases found associated with clinical aspergillosis in falcons,'' International Journal of Microbiology, vol

W. Tarello, ``Etiologic agents and diseases found associated with clinical aspergillosis in falcons,'' International Journal of Microbiology, vol. 2011, 2011

work page 2011
[7]

M \"u ller, J

M. M \"u ller, J. Kinne, R. Schuster, and J. Walochnik, ``Outbreak of microsporidiosis caused by enterocytozoon bieneusi in falcons,'' Veterinary parasitology, vol. 152, no. 1-2, pp. 67--78, 2008

work page 2008
[8]

M. A. Mohamed et al., ``Radiographic imaging, endoscopy and haematological profile as indispensable diagnostic tools in falcon health and disease,'' Ph.D. dissertation, Sudan University of Science and Technology, 2014

work page 2014

Show all 12 references

[9]

Sadeghi, A

M. Sadeghi, A. Banakar, S. Minaei, M. Orooji, A. Shoushtari, and G. Li, ``Early detection of avian diseases based on thermography and artificial intelligence,'' Animals, vol. 13, no. 14, p. 2348, 2023

2023
[10]

J. Zhou, A. H. Gandomi, F. Chen, and A. Holzinger, ``Evaluating the quality of machine learning explanations: A survey on methods and metrics,'' Electronics, vol. 10, no. 5, p. 593, 2021

2021
[11]

Kopka and P

H. Kopka and P. W. Daly, A Guide to , 3rd ed. 1em plus 0.5em minus 0.4em Harlow, England: Addison-Wesley, 1999

1999
[12]

11em plus .33em minus .07em 4000 4000 100 4000 4000 500 `\.=1000 = #1 \@IEEEnotcompsoconly \@IEEEcompsoconly #1 * [1] 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEauthorblockAstyle \@IEEEcompsocnotconfonly \@IEEEauthordefaulttextstyle \@IEEEcompsocnotconfonly \@I...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [2]

Panthakkan, S

A. Panthakkan, S. Anzar, and W. Mansoor, ``Unleashing the power of efficientnet-convnext concatenation for brain tumor classification,'' in 2023 15th Biomedical Engineering International Conference (BMEiCON). 1em plus 0.5em minus 0.4em IEEE, 2023, pp. 1--5

work page 2023

[2] [1]

Fischer, L

D. Fischer, L. Van Waeyenberghe, C. Cray, M. Gross, E. Usleber, F. Pasmans, A. Martel, and M. Lierz, ``Comparison of diagnostic tools for the detection of aspergillosis in blood samples of experimentally infected falcons,'' Avian Diseases, vol. 58, no. 4, pp. 587--598, 2014

work page 2014

[3] [3]

Muller, T

M. Muller, T. Mannil, and A. George, ``Study on the most common bacterial infections in falcons in the united arab emirates,'' in Proceedings of the 27th Annual AAV Conference, San Antonio, TX, USA, 2006, pp. 6--10

work page 2006

[4] [4]

M. A. A. Hemeiri, A. A. de la Torre, K. Mohteshamuddin, B. A. Degafa, and G. Ameni, ``Retrospective study on the health problems of falcons in al ain, united arab emirates,'' bioRxiv, pp. 2021--02, 2021

work page 2021

[5] [5]

A. F. Alrefaei, ``Molecular detection and genetic characterization of trichomonas gallinae in falcons in saudi arabia,'' Plos one, vol. 15, no. 10, p. e0241411, 2020

work page 2020

[6] [6]

Tarello, ``Etiologic agents and diseases found associated with clinical aspergillosis in falcons,'' International Journal of Microbiology, vol

W. Tarello, ``Etiologic agents and diseases found associated with clinical aspergillosis in falcons,'' International Journal of Microbiology, vol. 2011, 2011

work page 2011

[7] [7]

M \"u ller, J

M. M \"u ller, J. Kinne, R. Schuster, and J. Walochnik, ``Outbreak of microsporidiosis caused by enterocytozoon bieneusi in falcons,'' Veterinary parasitology, vol. 152, no. 1-2, pp. 67--78, 2008

work page 2008

[8] [8]

M. A. Mohamed et al., ``Radiographic imaging, endoscopy and haematological profile as indispensable diagnostic tools in falcon health and disease,'' Ph.D. dissertation, Sudan University of Science and Technology, 2014

work page 2014