Quality Assessment of Dual-Parallel Edge Deblocking Filter Architecture for HEVC/H.265

Round 1

Reviewer 1 Report

1. In the results and discussion of Part 8, the experimental analysis is not detailed enough. Please explain to the readers in detail to make the research more rigorous and detailed.

2. Table 1 and Table 2 headers are inconsistent with text analysis.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper is based on the previous work of the authors [38]. It shares same architecture with [38] while the authors conducted quality assessment of the outputs and analyzed the results. However, the goal of the proposed framework is not clearly stated in the paper. It seems to be neither quality assessment metric for deblocking filter, or optimization of decoder parameters based on quality assessment of outputs. The authors are required to show the novelty of the paper, as the manuscript shares major architecture with the published paper [38]. Followings are the additional comments of the overall manuscript.

 

 

1.     The introduction of the paper is not written properly. Rather than the conceptual backgrounds of HEVC and AVC, more description of the proposed deblocking filter framework should be included.

2.     In section 3, the authors stated SSIM is the best metric for quality assessment of reconstructed video. However, SSIM does not take account of temporal information. It can be stated that SSIM more reflects human visual system (HVS) characteristics compared to PSNR in static scenes, while it cannot be simply applied to temporal domain due to additional temporal characteristics of HVS. Also, original SSIM is calculated using only luminance. However, as the framework the authors are dealing with is colored video. Eq. (3) seems to be an extended version of SSIM in color domain, which a proper citation and additional descriptions are needed.

3.     The metric described in section 3 is never used in performance analysis part. In section 8, the authors have analyzed results using only PSNR, which it is described as inferior compared to SSIM in section 3. It is recommended to make use of modern state-of-the-art video quality assessment metrics if the authors are intended to better reflect human perceptual characteristics.

4.     In section 4.2, the boundary strength computation unit is proposed. However, the method for calculating the BS is not depicted. Also, the components described in Figure 3 are not properly explained. It is recommended for the authors to add explanation on each block and criteria.

5.     Minor typos and grammatical errors should be modified throughout the paper.

 

[38] Christopher, Prayline Rajabai, and Sivanantham Sathasivam. "Five-stage pipelined dual-edge deblocking filter architecture for H. 265 video codec." IEICE Electronics Express (2019): 16-20190500.

Author Response

Please see the attachment

Author Response File: Author Response.pdf