Neural Network-Based Estimation of Distortion Sensitivity for Image Quality Prediction

Bosse, Sebastian; Becker, Sören; Fisches, Zacharias V.; Samek, Wojciech; Wiegand, Thomas

doi:10.1109/icip.2018.8451261

Cited by 5 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such an aspect motivated the next step, which was training both networks to estimate the shifting parameter from distorted images (second group). For the hypothesis in which distorted images convey relevant information about visual quality parameters, our results in Table II: (a) confirm that degraded images seem to carry more relevant visual quality information than reference images, as previously stated in [Kang et al 2014;Bosse et al 2016;Bosse et al 2018;Bosse et al 2019]; and (b) suggest the use of different receptive fields improves visual quality prediction. The reasons for (a) and (b) are not clear so far, but we speculate that CNNs are exposed to larger regions of the Q c × Q p space during training phase when distorted images are network inputs.…”

Section: Resultssupporting

confidence: 82%

“…We believe that two non-mutually exclusive factors may explain this result: (a) the proposed MDN is useless or poorly designed; and (b) reference images do not have much to offer in terms of visual quality information to be exploited. An indication that the latter option could be true is found in [Kang et al 2014;Bosse et al 2016;Bosse et al 2018;Bosse et al 2019], where the authors argued that distorted images carry richer information than the reference A multi-stream dense network with different receptive fields to assess visual quality…”

Section: Resultsmentioning

confidence: 99%

“…Statistics of natural images are locally structured and highly non-stationary, so that perceived quality varies not only globally across different images, but also spatially within a given image [Bell and Sejnowski 1997;Ruderman 1994]. Considering the spatial variability of distortion perception, the parameter d can take a different value at each pixel position (x, y), i.e., d(x, y) [Bosse et al 2018;Bosse et al 2019], giving rise to a spatially variant version of paM SE:…”

Section: Adapted Psnrmentioning

confidence: 99%

“…Although MSE is commonly used as loss function in regressions, the mean absolute error (MAE) has proven less sensitive to outliers. The loss function adopted in [Bosse et al 2018] and [Bosse et al 2018] is an indirect regression of the shifting parameter (d) expressed in (5) and depicted in Fig. 2a.…”

Section: Adapted Psnrmentioning

confidence: 99%

See 3 more Smart Citations

A multi-stream dense network with different receptive fields to assess visual quality

Gonçalves¹,

Zampolo²,

Barros³

2019

Anais Do VII Symposium on Knowledge Discovery, Mining and Learning (KDMiLe 2019)

View full text Add to dashboard Cite

The prediction of visual quality is crucial in image and video systems. Image quality metrics based on the mean square error prevail in the field, due to their mathematical straightforwardness, even though they do not correlate well with the visual human perception. Latest achievements in the area support that the use of convolutional neural networks (CNN) to assess perceptual visual quality is a clear trend. Results in other applications, like blur detection and de-raining, indicate the combination of different receptive fields (i.e., convolutional kernels with different dimensions) improves a CNN performance. However, to the best of our knowledge, the role of different receptive fields in visual quality characterization is still an open issue. Thus, in this paper, we investigate the influence of using different receptive fields to predict image distortion. Specifically, we propose a multi-stream dense network that estimates a spatially-varying quality metric parameter from either reference or distorted images. The performance of the proposed method is compared with a competing state-of-the-art approach by using a public image database. Results show the proposed strategy outperforms the competing technique when the quality metric parameter is estimated from degraded images.

show abstract

Section: Resultssupporting

confidence: 82%

Section: Resultsmentioning

confidence: 99%

Section: Adapted Psnrmentioning

confidence: 99%

Section: Adapted Psnrmentioning

confidence: 99%

See 2 more Smart Citations

A multi-stream dense network with different receptive fields to assess visual quality

Gonçalves¹,

Zampolo²,

Barros³

2019

Anais Do VII Symposium on Knowledge Discovery, Mining and Learning (KDMiLe 2019)

View full text Add to dashboard Cite

show abstract

“…Afterwards, measuring the PCC and the Root Mean Squared Error(RMSE) between the MOS and the fitted scores will provide an estimation of prediction accuracy for the tested IQMs. In this experiment, we utilized a 4-parameter logistic function, as it is recently utilized in [70], which automatically guaranties the monotone behaviour of the fit function. Hereafter, the PCC measured before and after the logistic regression is referred as PCC NoFit and PCC Fitted respectively.…”

Section: Discussionmentioning

confidence: 99%

Comprehensive Performance Analysis of Objective Quality Metrics for Digital Holography

Ahar¹,

Birnbaum²,

Chlipała³

et al. 2020

Preprint

View full text Add to dashboard Cite

Objective quality assessment of digital holograms has proven to be a challenging task. While prediction of perceptual quality of the recorded 3D content from the holographic wavefield is an open problem; perceptual quality assessment from content after rendering, requires a time-consuming rendering step and a multitude of possible viewports. In this research, we use 96 Fourier holograms of the recently released HoloDB database to evaluate the performance of wellknown and state-of-the-art image quality metrics on digital holograms. We compare the reference holograms with their distorted versions: (i) before rendering on the real and imaginary parts of the quantized complex-wavefield, (ii) after converting Fourier to Fresnel holograms, (iii) after rendering, on the quantized amplitude of the reconstructed data, and (iv) after subsequently removing speckle noise using a Wiener filter. For every experimental track, the quality metric predictions are compared to the Mean Opinion Scores (MOS) gathered on a 2D screen, light field display and a holographic display. Additionally, a statistical analysis of the results and a discussion on the performance of the metrics are presented. The tests demonstrate that while for each test track a few quality metrics present a highly correlated performance compared to the multiple sets of available MOS, none of them demonstrates a consistently high-performance across all four test-tracks.

show abstract

Perceptually Optimized Bit-Allocation and Associated Distortion Measure for Block-Based Image or Video Coding

Helmrich

Bosse

Siekmann

et al. 2019

2019 Data Compression Conference (DCC)

View full text Add to dashboard Cite

Neural Network-Based Estimation of Distortion Sensitivity for Image Quality Prediction

Cited by 5 publications

References 21 publications

A multi-stream dense network with different receptive fields to assess visual quality

A multi-stream dense network with different receptive fields to assess visual quality

Comprehensive Performance Analysis of Objective Quality Metrics for Digital Holography

Perceptually Optimized Bit-Allocation and Associated Distortion Measure for Block-Based Image or Video Coding

Contact Info

Product

Resources

About