Content-Oriented Learned Image Compression

Li, Meng; Gao, Shangyin; Feng, Yihui; Shi, Yibo; Wang, Jing

doi:10.1007/978-3-031-19800-7_37

Cited by 7 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, the effectiveness of the periodic compensation loss in image encoding is demonstrated in Fig. 12, where the compression model follows the approach proposed in reference [13]. Overall, our proposed periodic compensation loss can effectively address the problem of checkerboard artifacts.…”

Section: Ablationsmentioning

confidence: 91%

See 1 more Smart Citation

High Visual-Fidelity Learned Video Compression

Li,

Shi,

Wang

et al. 2023

Proceedings of the 31st ACM International Conference on Multimedia

View full text Add to dashboard Cite

With the growing demand for video applications, many advanced learned video compression methods have been developed, outperforming traditional methods in terms of objective quality metrics such as PSNR. Existing methods primarily focus on objective quality but tend to overlook perceptual quality. Directly incorporating perceptual loss into a learned video compression framework is nontrivial and raises several perceptual quality issues that need to be addressed. In this paper, we investigated these issues in learned video compression and propose a novel High Visual-Fidelity Learned Video Compression framework (HVFVC). Specifically, we design a novel confidence-based feature reconstruction method to address the issue of poor reconstruction in newly-emerged regions, which significantly improves the visual quality of the reconstruction. Furthermore, we present a periodic compensation loss to mitigate the checkerboard artifacts related to deconvolution operation and optimization. Extensive experiments have shown that the proposed HVFVC achieves excellent perceptual quality, outperforming the latest VVC standard with only 50% required bitrate. CCS CONCEPTS• Computing methodologies → Computer vision tasks; Image compression.

show abstract

Section: Ablationsmentioning

confidence: 91%

“…Therefore, adversarial training is adopted to produce visual friendly frames. Specifically, we employ the relativistic average discriminator, which has been proven to be effective in producing photo-realistic image [13]. The generator loss L 𝑅𝑎 𝐷 and the discriminator loss L 𝑅𝑎 𝐺 are formulated as:…”

Section: Loss Functionmentioning

confidence: 99%