Hand Kinematic Model Construction Based on Tracking Landmarks

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper presents an interesting and technically sound proposal for the construction of a hierarchical kinematic model of the hand based on landmark tracking data obtained using MediaPipe and RGB-D sensors. Comments are provided below to strengthen the research:

1.  The introduction covers several relevant backgrounds, however it is recommended to expand the discussion on the specific challenges of hand kinematics (e.g. redundancy, singularities, anatomical variability) and how your approach addresses them compared to other recent work. Include more up-to-date references from the last 3 years, if available.
2. The theoretical approach is comprehensive and well developed. However, it is suggested to strengthen the methodological section with additional technical details: i) Specific parameters of the sensor used (model, resolution, sampling rate); ii) Details of the Unity environment and handheld model used; iii) Consider publication of key code or at least pseudocode to improve reproducibility.
3. The study lacks a quantitative experimental validation, it is recommended to include error metrics, comparison with existing methods or a test with real subjects would strengthen the contribution and allow evaluating the accuracy of the proposed model. Although the work is implemented in Unity and transformations and visualizations are shown, it is recommended to include a case study or functional test (e.g., gesture recognition, interaction with virtual objects) to illustrate the real impact of the model.
4. Conclusions should be reworded to highlight more clearly the contributions, limitations and possible future lines of research, such as integration with neural networks, prosthetics or rehabilitation environments.
5. The writing requires a thorough review of technical English and grammar. There are long sentences, with ambiguous or incorrect structures, which affect the clarity of the document. Professional proofreading is strongly recommended.

Author Response

Please view the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1. In the paper, it appears that a unified parameter (e.g., hand size, joint length, etc.) is applied when constructing a kinematic model of the hand. However, in reality, there are differences in hand size and joint range of motion between individuals. I am curious to know how these individual differences and limitations are handled in the paper. Additionally, if these differences are not reflected, it is anticipated that unrealistic hand postures may be reproduced, or user-specific parameter adjustments may be necessary. Could you please explain how you have considered these issues in the paper, or whether you plan to address them in the future?
2. In the paper, it appears that you have defined three length parameters and four angle parameters (joint angle, 0–90 degrees) for each finger. However, the first joint (MCP joint) of the actual finger can move upward from the horizontal position, but it seems that such movements (e.g., bending the finger backward) are not reflected in the angle parameters in the paper. I am curious if there is a specific reason why such joint extension movements were not included in the model, or if there are plans to incorporate them in the future.
3. The paper does not seem to provide a specific explanation of occlusion situations. For example, if the index finger or little finger of one hand is obscured from the camera's field of view by another finger, how would this affect subsequent tracking or kinematic modeling? I am also curious about how the model would operate or be corrected in such situations. If you have considered any correction methods or response strategies for occlusion, could you please explain them?
4. I have a question regarding the experimental results. Based on the results image in Fig. 13 alone, it seems difficult to judge how accurately the model reflects multiple overlapping fingers or various finger positions. I would like to know if there are any additional explanations or experiments or results that verify the accuracy of the model even with complex finger positions.
5. Looking at the result image in Fig. 14, it is difficult to visually confirm how different the actual hand shape is from the result because the hand posture on the left side is not clearly visible. In particular, there are parts of 14-(c) that look similar to 14-(d). I wonder if there are any clearer photos or results that show a clear difference even if the image quality is low.

Author Response

Please view the attached file.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript entitled "Hand Kinematic Model Construction Based on Tracking Landmarks" has been reviewed. While the technical content is strong, the manuscript suffers from several critical shortcomings in scientific presentation, clarity, and validation. The novelty of the proposed method is also not clearly distinguished from existing work. 

The study lacks an empirical evaluation of the proposed model’s accuracy or performance. No benchmark comparison, quantitative error analysis, or application demonstration is provided.

The novelty of this study is unclear. The authors should explicitly identify what methodological contributions are new.

The manuscript does not sufficiently discuss the state-of-the-art in MediaPipe-based kinematic modeling, including recent works that already integrate depth for 3D modeling. The review needs to be more comprehensive.

Comments on the Quality of English Language

The manuscript contains many grammatical and syntactic errors that hinder readability. Substantial editing is needed to improve clarity and flow.

Author Response

Please view the attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have revised the manuscript based on the comments, and it seems that the revised one can be accepted for publication.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript has been substantially improved. The authors replied satisfactorily to reviewers’ comments and took into account the suggestions given.