A Hybrid Quality Metric for Non-Integer Image Interpolation

Chen, Jinling; Xu, Yiwen; Ma, Kede; Huang, Huiwen; Zhao, Tiesong

doi:10.1109/qomex.2018.8463405

Cited by 8 publications

(9 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use two quality assessment metrics for image super-resolution, Energy Similarity (ES) [31] and Frequency Similarity (FS) [32], to measure the similarity between integer-and non-integer-scaling low-resolution images. ES can measure global visual information degradations between two patches.…”

Section: Optimal Integer Scaling-factor Selection Based On Similaritymentioning

confidence: 99%

See 1 more Smart Citation

Single Image Super-Resolution: From Discrete to Continuous Scale Without Retraining

et al. 2020

View full text Add to dashboard Cite

Convolutional neural network (CNN)-based single image super-resolution (SR) methods have achieved superior performance on some discrete-scaling factors, including 2, 3, and 4. However, the scaling factors for SR should be continuous and not discrete in practical applications. Previous CNN-based SR models usually yield poor results for non-integer-scaling factors and are sometimes even worse than results derived from the conventional bicubic method. To extend CNN-based SR models to continuous scale, this paper proposes a multiple-scaling-based SR (MSSR) method that combines an integer-scaling-factor SR and once or twice non-integer-scaling-factor SR without retraining networks. For a non-integer-scaling factor, the MSSR method first computes an optimal integer-scaling factor according to the data similarity and choose the corresponding pre-trained model for the next stage. Then, an existing CNN-based model is used to perform the integer-scaling-factor SR. Finally, the output is scaled to the target size. The proposed MSSR method can extend a variety of existing CNN-based SR models from discrete to continuous-scaling factors. Experimental results with six CNN-based SR models demonstrated that the MSSR method could effectively improve the performance of existing CNN-based SR models for continuous-scaling-factor SR without retraining networks. Furthermore, the comparison with a magnification-arbitrary method, called Meta-SR, shows that the proposed MSSR method usually outperforms Meta-SR for scaling factors greater than or equal to 2. INDEX TERMS Convolutional neural network, image interpolation, super-resolution.

show abstract

Section: Optimal Integer Scaling-factor Selection Based On Similaritymentioning

confidence: 99%

“…A smaller value of ES indicates better performance. According to [32], FS is computed in the Fourier domain as follows,…”

Section: Optimal Integer Scaling-factor Selection Based On Similaritymentioning

confidence: 99%

Single Image Super-Resolution: From Discrete to Continuous Scale Without Retraining

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Zhou et al proposed a Quality Assessment Database for Super-resolved images (QADS) and an IQA method considering the structural and textural components of images [20]. Chen et al presented a hybrid quality metric for non-integer image interpolation that combined both reduced-reference and no-reference philosophies [14]. Fang et al introduced a reduced-reference quality assessment method for image SR by predicting the energy and texture similarity between LR and HR images [16].…”

Section: Related Workmentioning

confidence: 99%

“…The performance evaluation results of the proposed DISQ model are listed in Table V and compared with other IQA methods, where the best and second-best results are shown in red and blue, respectively. The compared algorithms include four no-reference IQA methods (DIIVIVE [30], BRISQUE [31], HOSA [32], CNN-IQA [11]) designed for general distorted images, six no-reference ISA indicators (LPC-SI [36], MLV [37], GPC [38], SPARISH [39], Synthetic-MaxPol [40], HVS-MaxPol [35]), two focus quality assessment algorithms (FQPath [49], FocusLiteNN [50]), and four related SR-IQA works (NSS-SR [15], HYQM [14], LNQM [9], BSRIQA [45]).…”

Section: B Performance Comparisonmentioning

confidence: 99%

“…Compared with other data-driven models, CNN is capable of learning features and making quality predictors jointly in an end-to-end manner. Despite its superiority, CNN has not yet been well exploited in SR-IQA, for which most methods are not based on deep learning [9], [10], [14]- [16]. In this work, we propose a CNN-based SR-IQA algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Learning-Based Quality Assessment for Image Super-Resolution

Zhao¹,

Lin²,

Xu³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Image Super-Resolution (SR) techniques improve visual quality by enhancing the spatial resolution of images. Quality evaluation metrics play a critical role in comparing and optimizing SR algorithms, but current metrics achieve only limited success, largely due to the lack of large-scale quality databases, which are essential for learning accurate and robust SR quality metrics. In this work, we first build a largescale SR image database using a novel semi-automatic labeling approach, which allows us to label a large number of images with manageable human workload. The resulting SR Image quality database with Semi-Automatic Ratings (SISAR), so far the largest of SR-IQA database, contains 8,400 images of 100 natural scenes. We train an end-to-end Deep Image SR Quality (DISQ) model by employing two-stream Deep Neural Networks (DNNs) for feature extraction, followed by a feature fusion network for quality prediction. Experimental results demonstrate that the proposed method outperforms state-of-the-art metrics and achieves promising generalization performance in crossdatabase tests. The SISAR database and DISQ model will be made publicly available to facilitate reproducible research.

show abstract