Perceptual VVC quantization refinement with ensemble learning

Wu, Yuxuan; Wang, Zheng; Chen, Weiling; Lin, Liqun; Wei, Hongan; Zhao, Tiesong

doi:10.1016/j.displa.2021.102103

Cited by 7 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We adopt the single stimulus (SS) method to gather subjective ratings. Then, we properly integrate the existing FR IQA methods (Wu et al, 2021 ) to design an artificial intelligence (AI)-based gastroscope image quality evaluator (GIQE). To define it more concretely, we learn multiple kinds of HVS inspired features from gastroscope motion blurred images by the newly proposed semi-full combination subspace.…”

Section: Introductionmentioning

confidence: 99%

Subjective and objective quality assessment of gastrointestinal endoscopy images: From manual operation to artificial intelligence

Yuan

Bai²,

Yan³

et al. 2023

Front. Neurosci.

View full text Add to dashboard Cite

Gastrointestinal endoscopy has been identified as an important tool for cancer diagnosis and therapy, particularly for treating patients with early gastric cancer (EGC). It is well known that the quality of gastroscope images is a prerequisite for achieving a high detection rate of gastrointestinal lesions. Owing to manual operation of gastroscope detection, in practice, it possibly introduces motion blur and produces low-quality gastroscope images during the imaging process. Hence, the quality assessment of gastroscope images is the key process in the detection of gastrointestinal endoscopy. In this study, we first present a novel gastroscope image motion blur (GIMB) database that includes 1,050 images generated by imposing 15 distortion levels of motion blur on 70 lossless images and the associated subjective scores produced with the manual operation of 15 viewers. Then, we design a new artificial intelligence (AI)-based gastroscope image quality evaluator (GIQE) that leverages the newly proposed semi-full combination subspace to learn multiple kinds of human visual system (HVS) inspired features for providing objective quality scores. The results of experiments conducted on the GIMB database confirm that the proposed GIQE showed more effective performance compared with its state-of-the-art peers.

show abstract

Section: Introductionmentioning

confidence: 99%

Subjective and objective quality assessment of gastrointestinal endoscopy images: From manual operation to artificial intelligence

Yuan

Bai²,

Yan³

et al. 2023

Front. Neurosci.

View full text Add to dashboard Cite

show abstract

“…In recent years, machine-/deep-learning-based modeling of visual perception has become a new research trend due to the rapid development of deep neural networks [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ]. To fill the vacancy of JND databases for image and video compression, several scholars have suggested MCL-JCI [ 34 ], JND-pano [ 35 ], MCL-JCV [ 36 ], VideoSet [ 37 ], etc.…”

Section: Introductionmentioning

confidence: 99%

“…To fill the vacancy of JND databases for image and video compression, several scholars have suggested MCL-JCI [ 34 ], JND-pano [ 35 ], MCL-JCV [ 36 ], VideoSet [ 37 ], etc. Based on these datasets, various modeling techniques for JND eventually emerged, most notably subjective data regression [ 25 , 27 ], binary classification [ 26 ], picture/video-wise JND (PWJND/VWJND) or satisfied user ratio (SUR) modeling [ 27 , 28 , 29 , 30 ], and finding appropriate weighting factors for the JND models [ 25 , 26 ]. In particular, motivated by the smoothing performance (non-smoothed regions have better capability for hiding noise than smoothed regions), Wu et al [ 31 ] proposed an unsupervised learning method for generating JND images in the pixel domain.…”

Section: Introductionmentioning

confidence: 99%

Semantically Adaptive JND Modeling with Object-Wise Feature Characterization, Context Inhibition and Cross-Object Interaction

Wang

Yin

et al. 2023

Sensors

View full text Add to dashboard Cite

Performance bottlenecks in the optimization of JND modeling based on low-level manual visual feature metrics have emerged. High-level semantics bear a considerable impact on perceptual attention and subjective video quality, yet most existing JND models do not adequately account for this impact. This indicates that there is still much room and potential for performance optimization in semantic feature-based JND models. To address this status quo, this paper investigates the response of visual attention induced by heterogeneous semantic features with an eye on three aspects, i.e., object, context, and cross-object, to further improve the efficiency of JND models. On the object side, this paper first focuses on the main semantic features that affect visual attention, including semantic sensitivity, objective area and shape, and central bias. Following that, the coupling role of heterogeneous visual features with HVS perceptual properties are analyzed and quantified. Second, based on the reciprocity of objects and contexts, the contextual complexity is measured to gauge the inhibitory effect of contexts on visual attention. Third, cross-object interactions are dissected using the principle of bias competition, and a semantic attention model is constructed in conjunction with a model of attentional competition. Finally, to build an improved transform domain JND model, a weighting factor is used by fusing the semantic attention model with the basic spatial attention model. Extensive simulation results validate that the proposed JND profile is highly consistent with HVS and highly competitive among state-of-the-art models.

show abstract