SUM-GAN-GEA: Video Summarization Using GAN with Gaussian Distribution and External Attention

Round 1

Reviewer 1 Report

This work presents a solution for video summarization. 

The novelty of this work is incremental, building on published techniques. 

The paper needs to checked for errors, for example there are two figure 1, one of which should be figure 2, while figure 2 should be figure 7. Moreover, both equations (4) and (5) define P, where obviously one of them should be defining R. Furthermore, there are no gray boxes defining the input change module and the external attention mechanism module in Figure 1. 

It is not clear why a Gaussian distribution is selected. There doesn't seem to be a scientific backing to this claim in the paper. No experimentation is provided to show that this is true. The authors rely on "we believe" which is not a scientific arguement. 

The model seems to perform better given more samples during training. The average is simply the average performance of SumMe and TVSum. It would be interesting to see how the model bahaves on the mixed datasets. 

While disussing TVSum, the authors state "and outperforms SUM-IndLU by 1.5% on the SumMe dataset with a clear advantage." but this has nothing to do and the authors should maintian discussion on the TVSum dataset. Note that SUM-IndLU performs better than the proposed solution on the TVSum dataset. 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

_{The novelty of this paper is not clear. The difference between present work and previous Works should be highlighted.}

_{The author needs to change the abstract and focus more on problem domain. Before the paper contributions, the author could precisely include the need of developing the proposed method.}

 _{Are there other factors to compare modern methods, such as time execution?   Other meaning  how did the authors apply the Augmentation technique?}

_{The author could better explain how “Related works” is actually related to the current study. It is not clear to the reader how the manuscript is similar to or differs from these related works.} 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

·         The author needs to change the abstract and focus more on problem domain. Before the paper contributions, the author could precisely include the need of developing the proposed method.

·         The novelty of this paper is not clear. The difference between present work and previous Works should be highlighted.

·         The author could better explain how “Related works” is actually related to the current study. It is not clear to the reader how the manuscript is similar to or differs from these related works.

·         Some recent works should be added, such as : https://doi.org/10.1016/j.imavis.2021.104229; https://doi.org/10.3390/math10050733

·         Results need more explanations. Additional analysis is required at each experiment to show the main purpose.

·         Authors must develop the framework/architecture of the proposed methods

·         The author has used some mathematical notations. Make sure that all the parameters are described. And also check the mathematical notations.

·         How did the authors apply the Augmentation technique?

·         The manuscript is well-organized and properly formatted. The authors are suggested to have the paper revised to improve the language.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

The authors answered alll comments