Pith. sign in

REVIEW 4 major objections 5 minor 26 references

Enhancing Mathematics Learning for Hard-of-Hearing Students Through Real-Time Palestinian Sign Language Recognition: A New Dataset

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper introduces the first mathematical-gesture dataset for Palestinian Sign Language and reports a fine-tuned Vision Transformer at 97.59% accuracy.

desk verdict The new PSL math-gesture dataset is a genuine resource, but the paper's headline 97.59% accuracy is unsupported by the described split and should not be taken at face value. read the letter →

arxiv 2505.17055 v1 pith:ZD3NJC7V submitted 2025-05-16 cs.CL cs.CYcs.HC

Fidaa Khandaqji , Huthaifa I. Ashqar , Abdelrahem Atawnih This is my paper · ORCID
classification cs.CLcs.CYcs.HC
keywords PalestinianSignLanguagemathematicalgesturesVisionTransformerrecognitionassistivetechnologylow-resourceSTEMeducationcomputer
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 small, curated dataset can bring an under-served sign language into AI-driven education. It introduces 41 classes of mathematical gestures in Palestinian Sign Language, covering numbers, arithmetic operations, and geometric shapes, recorded by fluent signers and standardised against the official PSL dictionary. It then fine-tunes a Vision Transformer on frames from those videos and reports 97.59% accuracy, with precision and recall at the same level and F1-score at 97.58%. The sympathetic reading is that domain-specific datasets plus pretrained transformer backbones are enough to make real-time PSL math recognition feasible, reducing dependence on scarce human interpreters in Palestinian schools.

What carries the argument

The load-bearing mechanism is the Vision Transformer (ViT-base-patch16-224), which splits each 224x224 frame into 16x16 patches, adds positional embeddings, runs self-attention across patches, and finishes with an MLP head that outputs one of 41 gesture classes. It is fine-tuned from pretrained weights on 2,896 frames with horizontal flipping, rotation, color jitter, resizing, and normalization as augmentation. The dataset is the companion mechanism: 41 classes spanning numbers, operations, and shapes, each class represented by one video, with frames extracted to capture full sign movement; the paper states that PSL grammar, especially non-manual markers, guided the choice of which gestures to include.

What would settle it

Download the released 41-video dataset, reproduce the train/validation/test partition, and recompute accuracy after removing any frame whose source video appears in more than one split; if the 8 test classes are truly absent from training, a 41-class classifier cannot score above chance on them, so checking the class-to-video mapping and per-class predictions directly settles whether 97.59% is reproducible.

Watch

Extended reading notes

Core claim

The central claim is that a fine-tuned Vision Transformer can classify 41 mathematical signs of Palestinian Sign Language with a reported accuracy of 97.59%, precision of 97.59%, recall of 97.59%, and F1-score of 97.58%, using a new custom dataset of 2,896 frames. The discovery is the pairing of resource and model: the dataset is, by the authors' account, the first domain-specific PSL collection for mathematics, built from a standardised dictionary and recorded by 21 proficient signers, and the ViT is a general image-classification backbone rather than a bespoke sign-language architecture. The authors' claim is that this pairing closes a concrete gap, since PSL has been left out of the sign-language-recognition literature that focuses on ASL, ISL, and ArSL, and that the high metrics point to deployable AI tutoring for hard-of-hearing students in STEM classrooms.

Load-bearing premise

The 97.59% figure stands on the assumption that the test videos are genuine held-out samples of the 41 trained classes, with no frames from test videos leaking into training or validation, and that the reported split actually allows a 41-class classifier to be evaluated.

Editorial extensions

If this is right

  • Palestinian Sign Language gains a first domain-specific mathematical dataset and a reported classification baseline, giving later work something concrete to extend or contest.
  • The reported metrics support the feasibility of an AI layer for math education that responds to signed numbers, operations, and shapes in real time.
  • The fine-tuning recipe, pretrained transformer, small frame set, and standard augmentation, is light enough to be repeated for other low-resource sign languages with similar domain vocabularies.
  • If the dataset is widely used, future systems can combine this classifier with text-to-speech or on-screen math symbols, effectively translating PSL math signs into the written notation that hearing teachers and standard curricula expect.

Reading between the lines

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

  • The paper stops at classification; a natural extension the authors do not carry out is an end-to-end classroom test that measures latency, acceptance by deaf students, and whether real-time feedback improves math learning.
  • A static-frame ViT is a deliberate simplification; if temporal dynamics turn out to matter for PSL math signs, adding a recurrent or video-transformer layer should raise accuracy further, which is a testable comparison on the same dataset.
  • A signer-independent split, training and testing on disjoint sets of people, would be a stronger generalization test than the current video-level split, and is a concrete way to check whether the 97.59% accuracy holds beyond the 21 volunteers who recorded the data.
  • The paper's own future-work list, confusion matrices and per-class accuracy, is the minimal diagnostic for identifying confusable signs such as similar number hand shapes; until such numbers are reported, the aggregate 97.59% says little about which signs drive the result.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces a custom dataset of 41 Palestinian Sign Language (PSL) mathematical gesture classes, fine-tunes a ViT-base-patch16-224 model on it, and reports a classification accuracy of 97.59% along with similarly high precision, recall, and F1 scores. The authors claim this supports real-time mathematics-learning support for hard-of-hearing students and describe the dataset as the first domain-specific PSL mathematical-gesture resource. The methodology section describes data collection from 21 volunteers, a 24/9/8 video split into training, validation, and test sets, and a standard ViT fine-tuning pipeline.

Significance. If the empirical claims were sound, this paper would provide a useful first dataset for mathematical gestures in Palestinian Sign Language and demonstrate a practical classifier for an under-resourced sign language. The dataset idea and the focus on STEM education for deaf and hard-of-hearing students are genuinely valuable. However, the central performance claim is not supported by the described methodology: the data split as stated makes the reported accuracy impossible, and the manuscript provides no per-class results, no confusion matrix, and no released code. The contribution is therefore currently a dataset description with an invalid evaluation rather than a verified recognition result.

major comments (4)
  1. [§2.1, Table 2] The data split described in §2.1 is internally contradictory and invalidates the headline accuracy. The paper states that the dataset consists of 41 gesture videos, each representing a distinct class, and that the split is 24 training videos, 9 validation videos, and 8 test videos. If this is a class-level split, the 8 test classes are entirely absent from the training set, so a model trained on only 24 classes cannot correctly classify the 8 unseen test classes, and a reported 97.59% accuracy is arithmetically impossible. If the split is instead frame-level within videos, then frames from the same video appear in both training and test sets, which is data leakage and makes the reported metric an in-sample fit rather than an out-of-sample measurement. Either reading leaves the central accuracy claim unsupported.
  2. [§2.1 and Table 1] There are unexplained numerical inconsistencies in the reporting of the experimental setup. The text in §2.2 states a batch size of 23, while Table 1 lists a batch size of 32. The frame counts in §2.1 sum to 2316 + 289 + 290 = 2895, not the stated total of 2896. These discrepancies, together with the absence of released code, checkpoints, or per-class results, mean that the evaluation pipeline cannot be verified or reproduced from the manuscript.
  3. [Table 2 and §4] The paper reports only aggregate precision, recall, and F1 values and does not report the accuracy value in Table 2, despite the abstract claiming 97.59% accuracy. The conclusion in §4 explicitly lists confusion matrices and per-class accuracy as future work. Given the class-level split contradiction, the aggregate metrics alone cannot support the claimed effectiveness of the model on all 41 gesture classes.
  4. [Algorithm 1 and §2.2] The algorithm treats the model as processing frames independently: it loops over frames, assigns features from each frame, and then classifies using only the features after the loop. No temporal aggregation or sequence modeling is specified, which contradicts the paper's claim that the model captures temporal dependencies of dynamic PSL gestures. This is a conceptual gap for a system described as recognizing sign-language movement rather than isolated static frames.
minor comments (5)
  1. [Equations (1)–(5)] The typeset equations for accuracy, precision, recall, and F1 are missing their denominators, and the definition of TN repeats the definition of TP ('correctly predicted positive instances') instead of stating negative instances.
  2. [Abstract] The abstract contains the phrase 'inclusive frosting digital integration,' which appears to be a typographical error; it should likely be 'inclusive digital integration.'
  3. [Table 3 heading] The heading 'Compassion of sign language datasets' should read 'Comparison of sign language datasets.'
  4. [§2.2] There are minor typographical errors, including 'board compatibility' for 'broad compatibility,' 'tracnsformer_encoder' for 'transformer_encoder' in Algorithm 1, and 'Tabel 2' for 'Table 2.'
  5. [§2.1] The text describes the volunteers as '21 proficient volunteers in PSL' but the significance of the number of videos per class and the relationship between volunteers and gesture classes is not explained; providing per-class participant counts would clarify dataset diversity.

Circularity Check

1 steps flagged · score 6.0 of 10

The reported 97.59% test accuracy is not an independent measurement under the paper's own split description: video-level splitting makes test classes unseen, while frame-level splitting allows train/test overlap.

  1. fitted input called prediction [Section 2.1 (Data Collection and Preprocessing) and Section 3/Table 2 (Results)]
    "It consisted of 41 gesture videos, each representing a distinct class. ... 1. Training data: 24 videos 2316 frames, 80% of the dataset were used for Training ... 3. Testing data: Eight videos, 290 frames. Ten percent of the dataset was reserved for final evaluation, ensuring an unbiased assessment of the model's classification capabilities. ... All frames were extracted from the video and used in training, ensuring that the model captured the full sequence of movements for accurate sign recognition."

    Central claim is 97.59% test accuracy (Table 2), presented as unbiased evaluation. The split makes that impossible or leaky. If 41 videos are class-disjoint and split by video, 8 test classes are absent from 24 training classes, so 97.59% is arithmetically impossible. If the 80/10/10 split is frame-level (2316/2896, 289/2896, 290/2896), frames from the same video/class can appear in both train and test; 'All frames were extracted from the video and used in training' then makes test frames training-derived near-duplicates. Either way accuracy is forced by split construction, not an independent out-of-sample prediction.

full rationale

This is an empirical ML paper, not a formal derivation, so most circularity patterns (self-definition, uniqueness theorems, ansatz-by-citation) do not apply. Prior-work citations are contextual, and I found no load-bearing self-citation: the authors' own prior work is not invoked to justify the accuracy claim. The only candidate for circularity is the evaluation claim, and the paper's own text supports flagging it: the class/video arithmetic contradicts a valid disjoint test set, and the frame-level alternative creates train/test overlap. The conclusion even concedes the lack of per-class accuracy and confusion matrices ('refine model evaluation using detailed performance metrics (e.g., confusion matrices and per-class accuracy)'), so the single aggregate 97.59% number cannot be independently verified. Additional inconsistencies (batch size 23 in the text vs 32 in Table 1; frame counts summing to 2895 vs stated 2896) reinforce that the evaluation pipeline is unreliably reported. Score 6: the headline generalization 'prediction' reduces, under the frame-level reading, to recognition of training-derived frames, and under the video-level reading it is arithmetically impossible; either way it is not an independent result.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central empirical claim rests on standard transfer-learning assumptions and, crucially, on a data split that is described inconsistently. No invented theoretical entities are introduced. The main assumptions are frame-level sufficiency and independence of the test split.

free parameters (3)
  • Learning rate = 2e-5
    Chosen by hand for stable fine-tuning; no search reported.
  • Batch size = 32 (inconsistent with text value 23)
    Selected to balance training time and performance; inconsistent reporting.
  • Epochs = 10
    Chosen to stop training; no early stopping criterion reported.
assumptions (4)
  • domain assumption Individual video frames carry sufficient information to classify mathematical PSL gestures
    The system classifies each frame independently (Algorithm 1), ignoring temporal dynamics that are central to sign language. This assumption is load-bearing because dynamic signs may not be recognizable from static frames.
  • domain assumption The Palestinian Red Crescent PSL dictionary is an accurate and authoritative source for the recorded gestures
    The dataset is grounded in this dictionary, but no independent linguistic validation is reported.
  • domain assumption The data split yields a test set independent of training data
    Section 2.1 describes a video-level split, but the numbers imply either unseen classes or frame leakage; the derivation of accuracy depends on this assumption.
  • domain assumption ImageNet-pretrained ViT features transfer to PSL gesture frames
    The model uses ImageNet weights and the paper assumes these features are useful for sign language gestures; no ablation or comparison is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Mathematics Learning for Hard-of-Hearing Students Through Real-Time Palestinian Sign Language Recognition: A New Dataset." pith.science (2026). https://pith.science/paper/ZD3NJC7V

@misc{pith2026250517055,
  author       = {Pith},
  title        = {Pith review of: Enhancing Mathematics Learning for Hard-of-Hearing Students Through Real-Time Palestinian Sign Language Recognition: A New Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZD3NJC7V}},
  note         = {Machine review of arXiv:2505.17055}
}
read the original abstract

The study aims to enhance mathematics education accessibility for hard-of-hearing students by developing an accurate Palestinian sign language PSL recognition system using advanced artificial intelligence techniques. Due to the scarcity of digital resources for PSL, a custom dataset comprising 41 mathematical gesture classes was created, and recorded by PSL experts to ensure linguistic accuracy and domain specificity. To leverage state-of-the-art-computer vision techniques, a Vision Transformer ViTModel was fine-tuned for gesture classification. The model achieved an accuracy of 97.59%, demonstrating its effectiveness in recognizing mathematical signs with high precision and reliability. This study highlights the role of deep learning in developing intelligent educational tools that bridge the learning gap for hard-of-hearing students by providing AI-driven interactive solutions to enhance mathematical comprehension. This work represents a significant step toward innovative and inclusive frosting digital integration in specialized learning environments. The dataset is hosted on Hugging Face at https://huggingface.co/datasets/fidaakh/STEM_data.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 25 canonical work pages

  1. [1]

    https://www.who.int/ar/news-room/fact-sheets/detail/deafness-and-hearing-loss

    work page

  2. [2]

    https://www.un.org/ar/observances/sign-languages-day

    work page

  3. [3]

    H., Almasoudy, F

    Bashaa, M. H., Almasoudy, F. H., Sagheer, N. S., & Hadi, W. M. CNN Algorithm with SIFT to Enhance the Arabic Sign Language Recognition. International Journal of Emerging Science and Engineering, 12(10)

    work page

  4. [4]

    A., & Khamees, M

    Kadhim, R. A., & Khamees, M. (2020). A Real -Time American Sign Language Recognition System using Convolutional Neural Network for Real Datasets. Tem Journal, 9(3)

    work page 2020

  5. [5]

    Alawneh, K., & Abdel -Fattah, M. (2021). Deaf education in Palestine: Reality and Aspirations. BATOD Magazine

    work page 2021

  6. [6]

    H., Noor, A., Alharbi, A

    Noor, T. H., Noor, A., Alharbi, A. F., Faisal, A., Alrashidi, R., Alsaedi, A. S., ... & Alsaeedi, A. (2024). Real -Time Arabic Sign Language Recognition Using a Hybrid Deep Learning Model. Sensors, 24(11), 3683

    work page 2024

  7. [7]

    Real -time Arabic Sign Language Recognition based on YOLOv5

    Aiouez, S.; Hamitouche, A.; Belmadoui, M.S.; Belattar, K.; Souami, F. Real -time Arabic Sign Language Recognition based on YOLOv5. In Proceedings of the IMPROVE, Online Streaming, 22–24 April 2022; pp. 17–25

    work page 2022

  8. [8]

    Arabic sign language recognition using Faster R-CNN

    Alawwad, R.A.; Bchir, O.; Ismail, M.M.B. Arabic sign language recognition using Faster R-CNN. Int. J. Adv. Comput. Sci. Appl. 2021, 12, 692–700

    work page 2021

Show all 26 references
  1. [9]

    Egyptian sign language recognition using cnn and lstm

    Elhagry, A.; Elrayes, R.G. Egyptian sign language recognition using cnn and lstm. arXiv 2021, arXiv:2107.13647

    2021 arXiv

  2. [10]

    A Deep Learning based Approach for Recognition of Arabic Sign Language Letters

    Hdioud, B.; Tirari, M.E.H. A Deep Learning based Approach for Recognition of Arabic Sign Language Letters. Int. J. Adv. Comput. Sci. Appl. 2023, 14, 424–429

    2023

  3. [11]

    Spelling correction real -time American sign language alphabet translation system based on YOLO network and LSTM

    Rivera-Acosta, M.; Ruiz -Varela, J.M.; Ortega -Cisneros, S.; Rivera, J.; Parra -Michel, R.; Mejia-Alvarez, P. Spelling correction real -time American sign language alphabet translation system based on YOLO network and LSTM. Electronics 2021, 10, 1035

    2021

  4. [12]

    Sign Language Recognition Using Convolutional Neural Networks

    Uyyala, P. Sign Language Recognition Using Convolutional Neural Networks. J. Interdiscip. Cycle Res. 2022, 14, 1198–1207

    2022

  5. [13]

    Detection and interpretation of Indian Sign Language using LSTM networks

    Vyavahare, P.; Dhawale, S.; Takale, P.; Koli, V.; Kanawade, B.; Khonde, S. Detection and interpretation of Indian Sign Language using LSTM networks. J. Intell Syst. Control 2023, 2, 132–142

    2023

  6. [14]

    Indonesian sign language recognition using YOLO meod

    Daniels, S.; Suciati, N.; Fathichah, C. Indonesian sign language recognition using YOLO meod. In Proceedings of the IOP Conf. on Materials Science and Engineering, Yogyakarta, Indonesia, 13–14 November 2020; pp. 1–9

    2020

  7. [15]

    TMS -Net: A multi-feature multi- stream multi -level information sharing network for skeleton -based sign language recognition

    Deng, Z.; Leng, Y.; Chen, J.; Yu, X.; Zhang, Y.; Gao, Q. TMS -Net: A multi-feature multi- stream multi -level information sharing network for skeleton -based sign language recognition. Neurocomputing 2024, 572, 127194. 12

    2024

  8. [16]

    Natural language -assisted sign language recognition

    Zuo, R.; Wei, F.; Mak, B. Natural language -assisted sign language recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada, 18–22 June 2023; pp. 14890–14900

    2023

  9. [17]

    Sign language recognition with recurrent neural network using human keypoint detection

    Ko, S.K.; Son, J.G.; Jung, H. Sign language recognition with recurrent neural network using human keypoint detection. In Proceedings of the 2018 Conference on Research in Adaptive and Convergent Systems, Honolulu, HI, USA, 9–12 October 2018; pp. 326–328

    2018

  10. [18]

    Real-time Japanese sign language recognition based on three phonological elements of sign

    Sako, S.; Hatano, M.; Kitamura, T. Real-time Japanese sign language recognition based on three phonological elements of sign. In Proceedings of the HCI International 2016 – Posters’ Extended Abstracts: 18th International Conference, HCI International 2016, Toronto, ON, Canada,...

    2016

  11. [19]

    Real -time assamese sign language recognition using mediapipe and deep learning

    Bora, J.; Dehingia, S.; Boruah, A.; Chetia, A.A.; Gogoi, D. Real -time assamese sign language recognition using mediapipe and deep learning. Procedia Comput. Sci. 2023, 218, 1384–1393

    2023

  12. [20]

    Machine learning techniques for Indian sign language recognition

    Dutta, K.K.; Bellary, S.A.S. Machine learning techniques for Indian sign language recognition. In Proceedings of the 2017 International Conference on Current Trends in Computer, Electrical, Electronics, and Communication (CTCEEC), Mysore, India, 8 –9 September 2017; pp. 333–336

    2017

  13. [21]

    R., Bhatt, C

    Kothadiya, D. R., Bhatt, C. M., Saba, T., Rehman, A., & Bahaj, S. A. (2023). SIGNFORMER: deepvision transformer for sign language recognition. IEEE Access, 11, 4730-4739

    2023

  14. [22]

    F., & Maghari, A

    Alnabih, A. F., & Maghari, A. Y. (2024). Arabic Sign Language letters recognition using vision transformers. Multimedia Tools and Applications, 1-15

    2024

  15. [23]

    Sandoval-Castaneda, M., Li, Y., Brentari, D., Livescu, K., & Shakhnarovich, G. (2023). Self-supervised video transformers for isolated sign language recognition. arXiv preprint arXiv:2309.02450. Sign Language letters recognition using vision transformers. Multimedia Tools and ...

    2023 arXiv

  16. [24]

    A 3d -CLDNN based multiple data fusion framework for finger gesture recognition in human-robot interaction

    Qi, W.; Fan, H.; Xu, Y.; Su, H.; Aliverti, A. A 3d -CLDNN based multiple data fusion framework for finger gesture recognition in human-robot interaction. In Proceedings of the 2022 4th International Conference on Control and Robotics (ICCR), Guangzhou, China, 22–24 October 202...

    2022

  17. [25]

    Audio -visual speech and gesture recognition by sensors of mobile devices

    Ryumin, D.; Ivanko, D.; Ryumina, E. Audio -visual speech and gesture recognition by sensors of mobile devices. Sensors 2023, 23, 2284

    2023

  18. [26]

    Abdel-Fattah, M., & Alawnah, K. M. (2020). Modality in Palestinian Sign Language. International Journal of Innovation Creativity and Change

    2020

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.