Proceedings of the 1st Mile-High Video Conference 2022
DOI: 10.1145/3510450.3517289
|View full text |Cite
|
Sign up to set email alerts
|

Revisiting Bjontegaard delta bitrate (BD-BR) computation for codec compression efficiency comparison

Abstract: The Bjøntegaard Delta (BD) method proposed in 2001 has become a popular tool for comparing video codec compression efficiency. It was initially proposed to compute bitrate and quality differences between two Rate-Distortion curves using PSNR as distortion metric. Over the years, many works have calculated and reported BD results using other objective quality metrics such as SSIM, VMAF and, in some cases, even subjective ratings (mean opinion scores). However, the lack of consolidated literature explaining the … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
4
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
4
3
2
1

Relationship

0
10

Authors

Journals

citations
Cited by 12 publications
(5 citation statements)
references
References 37 publications
1
4
0
Order By: Relevance
“…Considering that, given a fixed bitrate, lower values of quality are associated to higher complexity of the content in terms of "compressibility", we consider the area under the quality-rate curve as an index of "compression complexity" or "compressibility". This is in line with the Bjontegaard delta method used to compare different codecs [26].…”
Section: Performance Indicatorssupporting
confidence: 81%
“…Considering that, given a fixed bitrate, lower values of quality are associated to higher complexity of the content in terms of "compressibility", we consider the area under the quality-rate curve as an index of "compression complexity" or "compressibility". This is in line with the Bjontegaard delta method used to compare different codecs [26].…”
Section: Performance Indicatorssupporting
confidence: 81%
“…This will also support an easy comparison of different compression methods based on SSIM. In fact, BD-rate and BDquality [11] were defined based on PSNR and their translation to different metrics such as SSIM is not obvious [12].…”
Section: Discussionmentioning
confidence: 99%
“…The experimental study evaluates the performance of deep learning-based video coding models, including STreamNet, DLIMD [18], and convLSTM [19], using five different datasets [20]: UCF10, HMDB51, Hollywood2, VID, and 20BN. The evaluation metrics include BPP, Bit Rate, PSNR [21], MS-SSIM [22], MSE, Compression Ratio, Encoding/Decoding Time, and BDBR [23]. Experiments use 10-fold cross-validation with 80% data for training, 10% for validation, and 10% for testing.…”
Section: ░ 4 Experimental Studymentioning
confidence: 99%