Transfer learning for just noticeable difference estimation

Mao, Yongwei; Wu, Jinjian; Wang, Xing; Li, Leida; Dong, Weisheng

doi:10.1016/j.ins.2023.119575

Cited by 2 publications

(1 citation statement)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[10] proposes a multitask deep learning framework to learn JND values, although without including the coding effects or evaluating its efficacy against the coding standards. [11] propose a pixel-wise JND model that uses transfer learning, but only for JPEG. [12], [13] proposed a linear regression based JND model for coding distortion to decide quantization parameters according to its entropy-based JND threshold.…”

Section: Introductionmentioning

confidence: 99%

IQNet: Image Quality Assessment Guided Just Noticeable Difference Prefiltering for Versatile Video Coding

Sun,

Lee,

Chang

2024

IEEE Open J. Circuits Syst.

View full text Add to dashboard Cite

Image prefiltering with just noticeable distortion (JND) improves coding efficiency in a visual lossless way by filtering the perceptually redundant information prior to compression. However, real JND cannot be well modeled with inaccurate masking equations in traditional approaches or imagelevel subject tests in deep learning approaches. Thus, this paper proposes a fine-grained JND prefiltering dataset guided by image quality assessment for accurate block-level JND modeling. The dataset is constructed from decoded images to include coding effects and is also perceptually enhanced with block overlap and edge preservation. Furthermore, based on this dataset, we propose a lightweight JND prefiltering network, IQNet, which can be applied directly to different quantization cases with the same model and only needs 3K parameters. The experimental results show that the proposed approach to Versatile Video Coding could yield maximum/average bitrate savings of 41%/15% and 53%/19% for all-intra and low-delay P configurations, respectively, with negligible subjective quality loss. Our method demonstrates higher perceptual quality and a model size that is an order of magnitude smaller than previous deep learning methods.

show abstract

Section: Introductionmentioning

confidence: 99%