Sign language recognition using the fusion of image and hand landmarks through multi-headed convolutional neural network
Pathan RK, Biswas M, Yasmin S, Khandaker MU, Salman M, Youssef AAF.
- DOI
- 10.1038/s41598-023-43852-x
- Record issued
- 2026-08-05
- Engine
- 7.15.0
- Exported
- 2026-09-21
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/b4f68698-50f8-43e0-84e0-84ec33d08402 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
- ReportingKey resources not met−0.5★
- LinksDead data/code link ×2−0.5★
- ReportingStudy design partially met−0.25★
- Statistics were not checked: no recomputable values were found in this text — no test statistic reported with its degrees of freedom, no effect estimate printed with both a 95% CI and a p-value, and no percentage printed with both its count and its denominator.
- 01Key resources not identified
The only applicable resource is software tools, which are named but not identified with versions or RRIDs, and custom code is not shared, resulting in a fail.
- 02Other integrity concern
Tortured phrase "convolutional neural organization" — an unnatural paraphrase of "convolutional neural network", a fingerprint of paraphrasing-tool / paper-mill text.
…the sign language gestures dataset on a convolutional neural organization model utilizing Inception v3. The model…
reviewer’s wording - 03Declared data/code link does not resolve
Dead link — nothing to verify.
“https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
- 04Declared data/code link does not resolve
Dead link — nothing to verify.
“https://www.kaggle.com/grassknoted/asl-alphabet”
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
The paper presents a sensible computational approach for ASL recognition with a multi-headed CNN, but has several reporting gaps: software tools lack version numbers, no code is shared, the data URL is dead, and there is no conflict of interest statement. The text contains grammatical errors and a tortured phrase ('convolutional neural organization') that raises concerns about potential language manipulation.
Three independent reviewer runs were synthesized. The statistical analysis dimension was not applicable because no inferential statistics were used. The copyedit pass flagged 30 issues (8 major). The verification components found 2 dead links and a tortured phrase indicating possible text generation.
Numerical inconsistencies
None foundValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
Checked — nothing surfaced.
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
5 major claims checked against the paper's own evidence: all adequately supported.
- partialReviewer 1The proposed method is suitable for 'wild situations' with complex backgrounds.The dataset used has complex backgrounds, which supports the claim. However, the method was not tested in truly uncontrolled environments (e.g., varying lighting, occlusions, multiple users), so the claim is somewhat overstated.Evidence: Dataset description: '...complex background with many users’ hand images to increase the detection complexity.' The paper does not test on a separate wild dataset.
“This method is also suitable for wild situations as it is not entirely dependent on hand position in an image frame.”
Introduction - partialReviewer 2The method is cost-effective and requires less computational power.The paper provides execution time and memory usage for the proposed model but does not compare these metrics with other state-of-the-art methods, making the claim of 'less computational power' only partially supported.Evidence: Table 5 shows execution time and memory for the three variants, but no external comparison.
Combined Model: Total Parameters 1,878,307 (7.17 MB), Execution time for 50 epochs 8230.36 seconds, Memory used 3030.80 MB.
Table 5reviewer’s wording - supportedReviewers 1, 2The proposed multi-headed CNN model achieves 98.98% test accuracy on the ASL Fingerspelling A dataset.The claim is directly supported by the results: Table 4 reports 98.98% accuracy, and the confusion matrix (Figure 7) and per-class metrics (Table 3) are consistent with this high accuracy.Evidence: Table 4 reports 'Proposed method: 98.98%' accuracy. Table 3 shows per-class precision, recall, and F1 scores all ≥0.95, supporting the overall accuracy.
“With the proposed model, 98.981% test accuracy has been achieved.”
Table 4 - supportedReviewer 1The fusion of image and hand landmarks improves recognition accuracy over single-channel approaches.The paper reports that the image-only model achieved 96.29% and the hand-landmark-only model achieved 98.42%, while the combined model reached 98.98%, demonstrating an improvement.Evidence: Result analysis: 'Before combining image and hand landmark inputs, we tested both individually and acquired a test accuracy of 96.29% for the image and 98.42% for hand landmarks.' Table 5 lists the combined model accuracy as 98.98%.
“Before combining image and hand landmark inputs, we tested both individually and acquired a test accuracy of 96.29% for the image and 98.42% for hand landmarks.”
Table 5 - supportedReviewer 2The fusion of image and hand landmarks improves accuracy over using either modality alone.Table 5 compares the combined model (98.98%) with image-only (96.29%) and hand-landmark-only (98.42%) models, showing the combined model achieves the highest accuracy.Evidence: Table 5 reports accuracies: Combined 98.98%, Image 96.29%, Hand Landmarks 98.42%.
Table 5 Complexity analysis of proposed model. Combined Model: Accuracy (%) 98.98; Image Model: 96.29; Hand landmarks model: 98.42.
Table 5reviewer’s wording
Data authenticity concerns
1 finding · worst mediumAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
- Other integrity concernAssessed
1 integrity concern flagged (0 high).
- mediumotherTortured phrase "convolutional neural organization" — an unnatural paraphrase of "convolutional neural network", a fingerprint of paraphrasing-tool / paper-mill text.
…the sign language gestures dataset on a convolutional neural organization model utilizing Inception v3. The model…
reviewer’s wording
Reporting gaps
2 findings · worst highRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Key resources not identifiedAssessed
- Study-design details incomplete (controls, blinding, power)Assessed
The paper reviews prior sign language recognition methods, noting their reliance on costly instruments, plain backgrounds, and high computational power. It then presents a clear rationale for using a multi-headed CNN with image and hand landmark fusion to address these issues. The limitations of prior work, such as dependence on single-channel CNN and plain backgrounds, are explicitly discussed and mitigated by the proposed approach.
“In this work, we have applied a fusion of traditional image processing with extracted hand landmarks and trained on a multi-headed CNN so that it could complement each other’s weights on the concatenation layer.”
“The literature summarizes that most of the models used in this application either depend on a single variable or require high computational power. Also, their dataset choice for training and validating the model is in plain background, which is easier to detect.”
“State-of-the-art techniques centered after utilizing deep learning models to improve good accuracy and less execution time.”
“The available online datasets are already processed or taken in a lab environment where it becomes easy for recent advanced AI models to train and evaluate, causing prone to errors in real-life applications with different kinds of noises.”
“However, such drawbacks can be easily overcome by employing artificial intelligence-based techniques. Since, in this modern era of advanced mobile technology, using a camera to take video or images is much easier, this study demonstrates a cost-effective technique to detect American Sign Language (ASL) using an image dataset.”
“The available online datasets are already processed or taken in a lab environment where it becomes easy for recent advanced AI models to train and evaluate, causing prone to errors in real-life applications with different kinds of noises. Accordingly, it is a basic need to make a model that can deal with noisy images and also be able to deliver positive results.”
“The literature summarizes that most of the models used in this application either depend on a single variable or require high computational power. Also, their dataset choice for training and validating the model is in plain background, which is easier to detect. Our main aim is to show how to reduce the computational power for training and the dependency of model training on one layer.”
For a computational study, the applicable sub-criteria are power_analysis (dataset selection), inclusion_exclusion (data curation), controls (baseline/ablations), outlier_handling, and independent_replication. The dataset is described with a 70-30 train-test split, but no formal power analysis or sample size justification is provided. Inclusion/exclusion criteria are clear (24 letters, excluding j and z). The paper presents ablation results comparing the combined model to image-only and hand-landmark-only models, serving as adequate controls. Outlier handling is not mentioned, and independent replication is not reported. Thus 2 of 5 applicable sub-criteria are adequate, yielding a warn.
“We intended to avoid that single color background and use a complex background with many users’ hand images to increase the detection complexity.”
“with 24 letters (except j and z as they contain motion)”
“Before combining image and hand landmark inputs, we tested both individually and acquired a test accuracy of 96.29% for the image and 98.42% for hand landmarks.”
“That’s why we have used the “ASL Finger Spelling” dataset , which has images of different sizes, orientations, and complex backgrounds of over 500 images per sign (24 sign total) of 4 users (non-native to sign language).”
“To avoid the overfitting problem, data augmentation and dynamic learning rate reduction have been used.”
“For training, we have used 46,023 images, and for testing, 19,725 images.”
The study uses a pre-existing image dataset for sign language recognition. While the dataset contains images of human hands, the study itself does not involve direct manipulation or collection of biological variables, making this dimension not applicable.
“over 500 images per sign (24 sign total) of 4 users (non-native to sign language).”
“this study demonstrates a cost-effective technique to detect American Sign Language (ASL) using an image dataset.”
The paper explicitly states the use of a publicly available dataset from Kaggle. This dataset consists of images and does not involve the collection of new human or animal subject data, nor does it use non-public identifiable data.
“The dataset used in this paper (ASL Fingerspelling Images (RGB & Depth)) is publicly available at Kaggle on this URL: https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
“The dataset used in this paper (ASL Fingerspelling Images (RGB & Depth)) is publicly available at Kaggle on this URL: https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
For a computational study, the applicable sub-criterion is software_tools_identified. The paper mentions Google's hand landmark model, ImageDataGenerator, Adam optimizer, and the Keras/TensorFlow framework, but does not provide version numbers, URLs, or RRIDs. No custom code is deposited in a public repository. Thus, the single applicable sub-criterion is reported_but_inadequate, leading to a fail.
“Google’s hand landmark model has an input channel of RGB and an image size of (224 × 224 × 3).”
“This model has been compiled with Adam optimizer and MSE loss for 50 epochs.”
“Image Data Generator did image augmentation with 10° rotation, 0.1 zoom range, 0.1 widths and height shift range, and horizontal flip.”
“The dataset used in this paper (ASL Fingerspelling Images (RGB & Depth)) is publicly available at Kaggle on this URL: https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
“Google’s hand landmark model has an input channel of RGB and an image size of (224 × 224 × 3).”
“This model has been compiled with Adam optimizer and MSE loss for 50 epochs.”
The paper reports test accuracy, per-class precision, recall, and F1 scores from a confusion matrix. No hypothesis tests, p-values, or effect sizes with confidence intervals are reported. The data presentation is adequate (confusion matrix, per-class metrics). Since no inferential statistics are used, the entire dimension is not applicable.
“For further evaluation, we have calculated the precision, recall, and F1 score of the proposed multi-headed CNN model, which shows excellent performance.”
“To compute these values, we first calculated the confusion matrix (shown in Fig. ).”
The paper includes a dedicated 'Data availability' section that explicitly states the dataset is publicly available and provides a direct URL to the Kaggle repository. No custom code is explicitly mentioned as being shared, but the dataset is the primary data resource.
“The dataset used in this paper (ASL Fingerspelling Images (RGB & Depth)) is publicly available at Kaggle on this URL: https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
“The dataset used in this paper (ASL Fingerspelling Images (RGB & Depth)) is publicly available at Kaggle on this URL: https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out”
Methods are described in sufficient detail for replication, including dataset preprocessing, model architecture, training parameters, and augmentation. All outcomes (accuracy, precision, recall, F1) are reported. Limitations are discussed in the conclusion. Conclusions are proportional to the evidence, with no overclaiming. Funding source is stated, but a conflicts-of-interest statement is absent. Five of six applicable sub-criteria are adequate, yielding a pass.
“Two-dimensional Convolutional layers with filter size 50, 25, kernel (3, 3) with Relu, strides 1; MaxPooling 2D with pool size (2, 2), batch normalization, and Dropout layer has been used in the hand landmarks training side.”
“Some limitations found in the proposed method compared with the literature. Some methods might work with low image dataset numbers, but as we use the simple CNN model, this method requires a good number of images for training.”
“Funding was provided by the American University of the Middle East, Egaila, Kuwait.”
“Two-dimensional Convolutional layers with filter size 50, 25, kernel (3, 3) with Relu, strides 1; MaxPooling 2D with pool size (2, 2), batch normalization, and Dropout layer has been used in the hand landmarks training side.”
“Some methods might work with low image dataset numbers, but as we use the simple CNN model, this method requires a good number of images for training.”
“There are some limitations found in the proposed method compared with the literature. Some methods might work with low image dataset numbers, but as we use the simple CNN model, this method requires a good number of images for training. Also, the proposed method depends on the hand landmark extraction model. Other hand landmark model can cause different results.”
“Funding was provided by the American University of the Middle East, Egaila, Kuwait.”
Broken references and links
1 finding · worst mediumReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
- Dead data/code linksRecomputed
Checked 32 references by DOI: 26 verified — 6 no DOI (shown, not verified).
- NO DOIKinect for Windows - Windows appsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOISign language recognition: State of the artNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAmerican sign language posture understanding with deep neural networksNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIDigital worlds that feel humanNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIInterpreting american sign language with KinectNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIASL AlphabetNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
2 data/code links checked; 0 live, 2 dead.
- datahttps://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-outDEADHTTP 404Dead link — nothing to verify.
- datahttps://www.kaggle.com/grassknoted/asl-alphabetDEADHTTP 404Dead link — nothing to verify.
Copyediting
1 finding · worst lowWording, consistency and formatting errors that need correcting before submission.
- Wording or formatting errors that need correctingAssessed
30 copyedit issues flagged (8 major): mostly grammar, clarity, typo.
- MAJORgrammarAbstract, first sentence“Sign Language Recognition is a breakthrough for communication among deaf-mute society”→ Sign Language Recognition is a breakthrough for communication within the deaf-mute communitySubject-verb agreement and article usage.
- MAJORclarityIntroduction, paragraph 2“Hard of hearing is a handicap that weakens their hearing and makes them unfit to hear, while quiet is an incapacity that impedes their talking and makes them incapable of talking.”→ Hard of hearing is a disability that impairs hearing, while muteness is a disability that impairs speech.Punctuation and word choice ('quiet' should be 'muteness').
- MAJORclarityPre-processing of image dataset, Raw image processing“In raw image processing, we have converted the images from RGB to grayscale to reduce color complexity.”→ In raw image processing, we converted the images from RGB to grayscale to reduce color complexity.Verb tense consistency.
- MAJORgrammarIntroduction, paragraph 1“Spoken language is the medium of communication between a majority of the population.”→ Spoken language is the primary medium of communication for the majority of the population.Awkward phrasing; 'between' is misused.
- MAJORgrammarIntroduction, paragraph 1“Mute people cannot convey a proper meaning using spoken language.”→ Mute people cannot convey meaning using spoken language.Unnecessary article 'a' and awkward wording.
- MAJORgrammarIntroduction, paragraph 1“Hard of hearing is a handicap that weakens their hearing and makes them unfit to hear, while quiet is an incapacity that impedes their talking and makes them incapable of talking.”→ Hearing impairment is a disability that weakens hearing and makes individuals unable to hear, while muteness is a disability that impedes speaking and makes individuals unable to talk.Inappropriate use of 'quiet' and 'handicap'; phrasing is unclear.
- MAJORtypoIntroduction, paragraph 3“Hand motions comprise a significant part of communication through signing vocabulary.”→ Hand motions comprise a significant part of sign language vocabulary.Awkward phrase 'through signing vocabulary'.
- MAJORgrammarConclusion, paragraph 1“It is highly implacable in real-world scenarios like communication, human–computer interaction, security, advanced AI, and much more.”→ It is highly applicable in real-world scenarios such as communication, human–computer interaction, security, advanced AI, and more.'Implacable' is likely a typo for 'applicable'.
- MINORtypoIntroduction, paragraph 2“Both are just handicapped in their hearing or potentially, therefore, cannot still do many other things.”→ Both are limited in their hearing or speech, but this does not prevent them from doing many other things.Awkward phrasing.
- MINORconsistencyResults, Table 1“Session | Total images per session | Depth | Intensity | Total images | | A | 12,547 | 0.49 pixel | 0.35 pixel | 65,748 |”→ Rows for sessions B-E missing depth and intensity data; clarify if only session A has these values.Table 1 is incomplete; depth and intensity columns are only filled for session A.
- MINORgrammarResult analysis, paragraph 4“We did not use binarization as it would affect the background of an image with skin color matched with hand color.”→ We did not use binarization as it would affect the background of an image with skin color that matches hand color.Awkward phrasing.
- MINORtypoConclusion, first sentence“This work proposes a methodology for perceiving the classification of sign language recognition.”→ This work proposes a methodology for classifying sign language recognition.Redundant phrase 'perceiving the classification'.
- MINORotherTable 4 caption“Table 4 Results of reviewed works for static image approaches.”→ Table 4 Comparison of reviewed works for static image approaches.Clarity.
- MINORpunctuationIntroduction, paragraph 2“images (2D, 3D), sensor data (hand globe , Kinect sensor , neuromorphic sensor ), videos, etc.”→ images (2D, 3D), sensor data (hand glove, Kinect sensor, neuromorphic sensor), videos, etc.Extra spaces before commas; 'hand globe' likely should be 'hand glove'.
- MINORclarityMethods, Working procedure“Before processing through a fully connected layer for classification, we merged both channel’s features so that the model could choose between the best weights.”→ Before passing through a fully connected layer for classification, we merged the features from both channels so that the model could select the best weights.Missing article and awkward phrasing.
- MINORgrammarConclusion, paragraph 1“For a long time, researchers have been working in this field to make a reliable, low cost and publicly available SRL system using different sensors, images, videos, and many more techniques.”→ For a long time, researchers have been working in this field to develop a reliable, low-cost, and publicly available sign language recognition system using different sensors, images, videos, and other techniques.Missing hyphen in 'low cost'; 'SRL' should be spelled out; 'many more' is vague.
- MINORgrammarAbstract“Although some of the previous studies have successfully recognized sign language, it requires many costly instruments including sensors, devices, and high-end processing power. However, such drawbacks can be easily overcome by employing artificial intelligence-based techniques.”→ Although some previous studies have successfully recognized sign language, they often require costly instruments, including sensors, devices, and high-end processing power. However, such drawbacks can be easily overcome by employing artificial intelligence-based techniques.Improved flow and subject-verb agreement.
- MINORclarityAbstract“Here, “Finger Spelling, A” dataset has been used, with 24 letters (except j and z as they contain motion).”→ Here, the “Finger Spelling, A” dataset has been used, comprising 24 letters (excluding 'j' and 'z' due to their motion-based nature).Clarified dataset name and reason for exclusion.
- MINORgrammarIntroduction, paragraph 1“Nonetheless, despite spoken language, a section of the population cannot speak with most of the other population.”→ Nonetheless, despite spoken language, a section of the population cannot communicate with the majority of the other population.Improved phrasing for clarity.
- MINORgrammarIntroduction, paragraph 1“Both are just handicapped in their hearing or potentially, therefore, cannot still do many other things.”→ Both are handicapped in their hearing or potentially, therefore, cannot do many other things.Removed redundant 'just' and 'still'.
- MINORclarityIntroduction, paragraph 2“All things are considered due to the fact that the captured images are excessively noisy. Therefore an elevated level of pre-processing is required.”→ All these factors are considered because captured images are often excessively noisy, thus requiring an elevated level of pre-processing.Improved sentence structure and flow.
- MINORgrammarIntroduction, paragraph 3“Sign language guides this part of the community and empowers smooth communication in the community of people with trouble talking and hearing (deaf and dumb).”→ Sign language guides this community and empowers smooth communication among people with hearing and speech impairments.Used more appropriate and respectful terminology.
- MINORclarityIntroduction, paragraph 3“Hand motions comprise a significant part of communication through signing vocabulary. At the same time, facial expressions and body activities assume the jobs of underlining the words and phrases communicated by hand motions.”→ Hand motions comprise a significant part of sign language vocabulary, while facial expressions and body movements serve to emphasize the words and phrases communicated by hand motions.Improved word choice ('movements' instead of 'activities', 'emphasize' instead of 'assume the jobs of underlining').
- MINORgrammarLiterature review, paragraph 1“State-of-the-art techniques centered after utilizing deep learning models to improve good accuracy and less execution time.”→ State-of-the-art techniques have focused on utilizing deep learning models to improve accuracy and reduce execution time.Improved verb tense and phrasing.
- MINORclarityLiterature review, paragraph 1“This is halfway because of the trouble in adjusting the CNNs to join both spatial and fleeting data.”→ This is partly due to the difficulty in adapting CNNs to combine both spatial and temporal data.Used more common terminology ('partly' instead of 'halfway', 'temporal' instead of 'fleeting').
- MINORclarityLiterature review, paragraph 2“But it is not sure it works with hand gestures as they detected face first and thenody movement .”→ However, it is uncertain if this approach works effectively with hand gestures, as it primarily detects faces and then body movement.Improved clarity and grammar.
- MINORgrammarModel building, paragraph 1“Before this, we trained processed images and hand landmarks with two separate models to compare.”→ Before this, we trained processed images and hand landmarks using two separate models for comparison.Improved phrasing.
- MINORgrammarTraining and testing, paragraph 1“For 50 epochs, the training vs testing accuracy and loss has been shown in Fig. .”→ For 50 epochs, the training versus testing accuracy and loss are shown in Fig. .Corrected verb tense.
- MINORclarityTraining and testing, paragraph 2“Numerical values in X and Y axis means the sequential letters from A = 0 to Y = 24, number 9 and 25 is missing because dataset does not have letter J and Z.”→ Numerical values on the X and Y axes represent sequential letters from A=0 to Y=24. Numbers 9 and 25 are missing because the dataset does not include letters 'J' and 'Z'.Improved clarity and punctuation.
- MINORgrammarResult analysis, paragraph 1“In human action recognition tasks, sign language has an extra advantage as it can be used to communicate efficiently.”→ In human action recognition tasks, sign language offers an additional advantage as it can be used to communicate efficiently.Improved word choice.
As a published paper, an informed reader should weigh the following: (1) the data availability link is dead, undermining reproducibility; (2) software versions are not reported, making the computational environment irreproducible; (3) the text contains a tortured phrase ('convolutional neural organization') that is a fingerprint of paraphrasing tools, warranting scrutiny of the manuscript's originality. These issues would warrant an erratum to fix the URL and add a COI statement, and ideally a public code repository for re-analysis.
- 1.HIGHdata codeFix the dead data availability link (currently https://www.kaggle.com/datasets/mrgeislinger/asl-rgb-depth-fingerspelling-spelling-it-out) or provide an alternative permanent repository (e.g., Zenodo) with a DOI.The reproducibility verification found both links dead; readers cannot access the dataset without a working link.
- 2.HIGHotherInvestigate and replace the tortured phrase 'convolutional neural organization' (which appears in the text as a paraphrase of 'convolutional neural network') with the correct term, and review the entire manuscript for similar unnatural paraphrases that may indicate text generation or paper mill activity.Such phrases are a known fingerprint of paraphrasing tools and undermine the integrity of the manuscript.
- 3.HIGHcopyeditRewrite the Abstract and Introduction for clarity and correct grammar, including fixing the misuse of 'quiet' for 'muteness', 'handicapped' for 'disabled', and 'implacable' (should be 'applicable') in the Conclusion.These errors significantly affect readability and professional presentation; a reader may question the rigor of the work if the writing is poor.
- 4.HIGHreportingAdd a conflict of interest statement in the Declarations section (e.g., 'The authors declare no competing interests.') and ensure it is compliant with the journal's requirements.This is a standard reporting requirement; its absence is a minor but noticeable gap.
- 5.HIGHdata codeDeposit the custom code (multi-headed CNN model, training, and evaluation scripts) in a public repository (e.g., GitHub, Zenodo) and cite it in the Data Availability section.Code sharing is essential for reproducibility of computational studies; its absence is a major gap.
- 6.MEDIUMrigorAdd version numbers and source URLs for all software tools used (e.g., Python, TensorFlow/Keras, OpenCV, Google MediaPipe, scikit-learn) in the Methods section.Without version information, the computational environment is not reproducible.
- 7.MEDIUMrigorAdd a statement on whether any power analysis or sample size justification was performed, or justify why none is needed for this benchmark study.A formal power analysis is not always required for computational benchmarks, but its absence should be acknowledged.
- 8.MEDIUMreportingComplete Table 1 by filling in the missing depth and intensity data for sessions B-E, or clarify that these values are only available for session A.An incomplete table undermines the transparency of the dataset description.
- 9.MEDIUMreportingAdd a statement on outlier handling (e.g., 'No outliers were excluded from the dataset') in the Methods section.Readers need to know whether any data were excluded and why.
- 10.MEDIUMcopyeditFix the awkward phrasing in the Abstract: 'it requires many costly instruments' should be 'they require many costly instruments' (subject-verb agreement with 'studies').Grammatical errors reduce the professionalism of the paper.
- 11.LOWreportingAdd a section on data curation detailing any inclusion/exclusion criteria for individual images (e.g., blurry images, unclear hands) to improve study design transparency.This would strengthen the description of dataset preparation.
- 12.LOWstatisticsIf the exact test set accuracy is 98.98%, report the number of correctly classified images out of the total test set (e.g., '19,525 out of 19,725') to allow verification.This provides a simple check for the reported accuracy.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.