Next Article in Journal
Coverage Path Planning Based on the Optimization Strategy of Multiple Solar Powered Unmanned Aerial Vehicles
Next Article in Special Issue
Persistent Charging System for Crazyflie Platform
Previous Article in Journal
Multimodal Fusion of Voice and Gesture Data for UAV Control
Previous Article in Special Issue
Rotor Failure Compensation in a Biplane Quadrotor Based on Virtual Deflection
 
 
Article
Peer-Review Record

The HDIN Dataset: A Real-World Indoor UAV Dataset with Multi-Task Labels for Visual-Based Navigation

by Yingxiu Chang 1, Yongqiang Cheng 1, John Murray 2,*, Shi Huang 3 and Guangyi Shi 4
Reviewer 1:
Reviewer 2:
Submission received: 28 June 2022 / Revised: 5 August 2022 / Accepted: 7 August 2022 / Published: 11 August 2022

Round 1

Reviewer 1 Report

The paper introduced a new indoor UAV dataset, Hull Drone Indoor Navigation (HDIN) dataset, for the visual-based navigation system. The dataset can be used for machine learning-based UAV control under real-world settings through regression and classification algorithms. The paper is well-written and well-structured. It has clear research motivation and good problem-solving strategies. The data collection and optimization approaches were introduced and discussed in good detail. In general, I would recommend the journal accept the paper after some minor amendments:

1. The research problem and the objectives should be clear at the beginning of the paper. It was until reading the second-page audience started to realize that the research work was to create “Navigation-Oriented datasets.”

2. The experiment is in three different buildings of the university. It is unclear why those scenarios were chosen (easy to deploy? Commonly used? or very challenging?). The samples are shown in the paper look well-lighted without moving objects like humans. Any consideration of using UAVs to work in some dangerous environments? Also, why can we not include more complicated scenarios such as T-shape crosses?

3. It may be helpful to show the quality of the dataset through the correlations and distributions analysis

4. It is unclear what regression and classification algorithms were used in section 4. 

5. In Figure 1(b), it is better to add some annotations to the logos (Ubuntu? WIFI?)\

6. In Figure 6, the resolutions of the plots from the first columns were not as good as in other columns

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have completely misplaced the Table 1 and the paragraph preceding that Table. These do not belong to the Introduction Section. The Introduction needs a complete overhaul and other relevant literature studied and described. Therea re also many typos and grammar mistakes throughout the paper. The "kind of challenges" stated in last sentence of first paragraph , has not been defined. The rest of the paper is written well, but needs to be re-evaluated after a through rewriting of the Introduction section.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

While the Introduction is now somewhat improved and more relevant, much more detailed discussion with other recent works also expected. The Introduction is still too flippant and short. 

2. Lines numbers 325-332 are replete with grammar mistakes. It is not clear what the authors are saying here. 

3. The algorithms 1-3 are part of the same dynamic process. It is suggested to not name them differently, but under the same algorithm with different parts. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Back to TopTop