No-Reference Image Quality Assessment Based on Multi-scale Convolutional Neural Networks

Chen, Peikun; Niu, Yuzhen; Huang, Dong

doi:10.1007/978-3-030-22868-2_84

“…Inspired by the work of Chen et al [32], we adopt a multiscale architecture for our convolutional neural network model, referred to as MSCNN. This technique aims to solve a problem with previous random patch methods, by paying equal attention to global and local features.…”

Section: Multi-scale Cnnmentioning

confidence: 99%

Portrait Quality Assessment using Multi-Scale CNN

Nicolas¹,

Belkarfa²

2021

lim

View full text Add to dashboard Cite

In this paper, we propose a novel and standardized approach to the problem of camera-quality assessment on portrait scenes. Our goal is to evaluate the capacity of smartphone front cameras to preserve texture details on faces. We introduce a new portrait setup and an automated texture measurement. The setup includes two custom-built lifelike mannequin heads, shot in a controlled lab environment. The automated texture measurement includes a Region-of-interest (ROI) detection and a deep neural network. To this aim, we create a realistic mannequins database, which contains images from different cameras, shot in several lighting conditions. The ground-truth is based on a novel pairwise comparison technology where the scores are generated in terms of Just-Noticeable-differences (JND). In terms of methodology, we propose a Multi-Scale CNN architecture with random crop augmentation, to overcome overfitting and to get a low-level feature extraction. We validate our approach by comparing its performance with several baselines inspired by the Image Quality Assessment (IQA) literature.

show abstract

“…Li et al [54] adopted multi-scale convolution to process image for super-resolution and obtained more effective high-frequency details, thus achieving good results. Inspired by this, Chen et al [37] also applied multi-scale feature extraction to the field of image quality assessment. Experiments show that multi-scale feature extraction can also improve the performance of quality assessment when processing images.…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, in order to solve the above problems, this study proposed a multi-scale residual CNN with attention mechanism to extract contour features and process detail features and channel correlation. Firstly, the multi-scale feature extraction module used is different from the above [54] and [37]. Multi-scale residual block (MSRB) with three parallel branches is used for feature extraction of distorted images.…”

Section: Introductionmentioning

confidence: 99%

No-reference image quality assessment with multi-scale weighted residuals and channel attention mechanism

Wang

¹

,

Lv

²

,

Ding

³

et al. 2022

View full text Add to dashboard Cite

Image is an important information source for human perception and machine pattern recognition. Image quality determines the accuracy and sufficiency of the information obtained. With the rapid development of deep learning in image processing, no-reference image quality assessment (NR-IQA) plays a significant role. Currently, most NR-IQA methods mainly use the global features of images without paying attention to the detail-rich local features and the dependencies between channels. There are subtle differences in detail between distorted and reference images, and the contribution of different channels to image quality assessment differs. Additionally, multi-scale feature extraction can be used to fuse the detailed information of images with different resolutions. The combination of global and local features plays a vital role in extracting image features. Therefore, a multi-scale residual convolutional neural network with an attention mechanism (MsRCANet) is proposed for NR-IQA. First, the network extracts global features and processes local features. Specifically, a multi-scale residual block is used to extract features from distorted images. Then, the residual learning with active weighted mapping strategy and channel attention mechanism is used to further process image features to obtain more abundant high-frequency information. Finally, the fusion strategy and full connection layer are used to evaluate the quality of the proposed network. The experimental results on four synthetic databases and three in-the-wild IQA databases, as well as validation results on cross-database, show that the proposed method has good generalization ability and can be compared with the most advanced methods.

show abstract

No-reference image quality assessment with multi-scale weighted residuals and channel attention mechanism

Wang

¹

,

Lv

²

,

Ding

³

et al. 2022

Preprint

0

View full text Add to dashboard Cite

Image is an important information source for human perception and machine pattern recognition. Image quality determines the accuracy and sufficiency of the information obtained. With the rapid development of deep learning in image processing, no-reference image quality assessment (NR-IQA) plays a significant role. Currently, most NR-IQA methods mainly use the global features of images without paying attention to the detail-rich local features and the dependencies between channels. There are subtle differences in detail between distorted and reference images, and the contribution of different channels to image quality assessment differs. Additionally, multi-scale feature extraction can be used to fuse the detailed information of images with different resolutions. The combination of global and local features plays a vital role in extracting image features. Therefore, a multi-scale residual convolutional neural network with an attention mechanism (MsRCANet) is proposed for NR-IQA. First, the network extracts global features and processes local features. Specifically, a multi-scale residual block is used to extract features from distorted images. Then, the residual learning with active weighted mapping strategy and channel attention mechanism is used to further process image features to obtain more abundant high-frequency information. Finally, the fusion strategy and full connection layer are used to evaluate the quality of the proposed network. The experimental results on four synthetic databases and three in-the-wild IQA databases, as well as validation results on cross-database, show that the proposed method has good generalization ability and can be compared with the most advanced methods.

show abstract

No-Reference Image Quality Assessment Based on Multi-scale Convolutional Neural Networks

Cited by 3 publications

References 32 publications

Portrait Quality Assessment using Multi-Scale CNN

Portrait Quality Assessment using Multi-Scale CNN

No-reference image quality assessment with multi-scale weighted residuals and channel attention mechanism

No-reference image quality assessment with multi-scale weighted residuals and channel attention mechanism

Contact Info

Product

Resources

About