Reduced reference IQA based on structural dissimilarity

Bhateja, Vikrant; Kalsi, Aseem; Srivastava, Aastha

doi:10.1109/spin.2014.6776923

Cited by 14 publications

(2 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…KLD being asymmetric is not suitable for IQA as the visual quality distance from one image to another should be identical. Further, its performance was consistent for certain distortion types only [37]- [40]. Estimation of DNT prior to the computation of KLD served to improve the performance of WNISM.…”

Section: A Motivationmentioning

confidence: 82%

A Reduced Reference Distortion Measure for Performance Improvement of Smart Cameras

et al. 2015

Self Cite

View full text Add to dashboard Cite

Application specific information processing (ASIP) unit in smart cameras requires sophisticated image processing algorithms for image quality improvement and extraction of relevant features for image understanding and machine vision. The improvement in performance as well as robustness can be achieved by intelligent moderation of the parameters both at algorithm (image resolution, contrast, compression, and so on) as well as hardware levels (camera orientation, field of view, and so on). This paper discusses the employment of ISO/IEC/IEEE 21451 smart transducer standards for performance improvement of smart cameras. The standardized transducer electronic data sheets (TEDS-by IEEE 21450) provide the self description of sensors, of which the calibration details are of vital importance to yield a smart and reconfigurable imaging system. This is possible by exercising intelligent control over the TEDS (smart camera) calibration details as well as automated tuning of algorithm parameters (in ASIP) based on decisions by perceptually efficient image quality assessment (IQA) tool. Estimation of distortion based on reduced reference IQA has been highlighted as a reliable methodology for this purpose. The proposed IQA approach uses wavelets for features extraction followed by estimation of luminance, contrast, and divergence parameters to obtain the proposed distortion measure (Q). The computational complexity in the process has been catalyzed using integral image and gradient magnitude approaches. The validation of Q metric is carried out by evaluating the image quality for various types of distortions on images from Content-based Strategies of Image Quality assessment (CSIQ) and Information Visualization CyberInfrastructure (IVC) databases. Simulation results yield a healthy correlation of Q and the subjective human opinions.

show abstract

Section: A Motivationmentioning

confidence: 82%

A Reduced Reference Distortion Measure for Performance Improvement of Smart Cameras

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition, existing study [75] suggests that feature dissimilarity among different scales is related to technical quality. As the dissimilarity is also noticed in our training process (Appendix Sec.…”

Section: Aesthetic Quality Evaluator (Aqe)mentioning

confidence: 99%

Disentangling Aesthetic and Technical Effects for Video Quality Assessment of User Generated Content

Wu¹,

Liao²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

User-generated-content (UGC) videos have dominated the Internet during recent years. While it is well-recognized that the perceptual quality of these videos can be affected by diverse factors, few existing methods explicitly explore the effects of different factors in video quality assessment (VQA) for UGC videos, i.e. the UGC-VQA problem. In this work, we make the first attempt to disentangle the effects of aesthetic quality issues and technical quality issues risen by the complicated video generation processes in the UGC-VQA problem. To overcome the absence of respective supervisions during disentanglement, we propose the Limited View Biased Supervisions (LVBS) scheme where two separate evaluators are trained with decomposed views specifically designed for each issue. Composed of an Aesthetic Quality Evaluator (AQE) and a Technical Quality Evaluator (TQE) under the LVBS scheme, the proposed Disentangled Objective Video Quality Evaluator (DOVER) reach excellent performance (0.91 SRCC for KoNViD-1k, 0.89 SRCC for LSVQ, 0.88 SRCC for YouTube-UGC) in the UGC-VQA problem. More importantly, our blind subjective studies prove that the separate evaluators in DOVER can effectively match human perception on respective disentangled quality issues. Codes and demos are released in https://github.com/teowu/dover.

show abstract