No-Reference Screen Content Image Quality Assessment With Unsupervised Domain Adaptation

Chen, Baoliang; Li, Haoliang; Fan, Hongfei; Wang, Shiqi

doi:10.1109/tip.2021.3084750

Cited by 28 publications

(13 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With the same purpose under a unified evaluation system, i.e., HVS, using DA is a plausible method to solve the problem of no-reference PCQA. According to the best of our knowledge, the only application of DA on quality assessment is [6], in which the authors used MMD as a loss function to predict screen content image quality via treating natural images as source domain. In this paper, we discuss the application of DA between 2D and 3D perception, which is more complex than the case in [6].…”

Section: Related Workmentioning

confidence: 99%

“…According to the best of our knowledge, the only application of DA on quality assessment is [6], in which the authors used MMD as a loss function to predict screen content image quality via treating natural images as source domain. In this paper, we discuss the application of DA between 2D and 3D perception, which is more complex than the case in [6]. According to the best of our knowledge, there is no study on PCQA via DA.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

No-Reference Point Cloud Quality Assessment via Domain Adaptation

Yang¹,

Liu²,

Chen³

et al. 2021

Preprint

View full text Add to dashboard Cite

We present a novel no-reference quality assessment metric, the image transferred point cloud quality assessment (IT-PCQA), for 3D point clouds. For quality assessment, deep neural network (DNN) has shown compelling performance on no-reference metric design. However, the most challenging issue for no-reference PCQA is that we lack large-scale subjective databases to drive robust networks. Our motivation is that the human visual system (HVS) is the decision-maker regardless of the type of media for quality assessment. Leveraging the rich subjective scores of the natural images, we can quest the evaluation criteria of human perception via DNN and transfer the capability of prediction to 3D point clouds. In particular, we treat natural images as the source domain and point clouds as the target domain, and infer point cloud quality via unsupervised adversarial domain adaptation. To extract effective latent features and minimize the domain discrepancy, we propose a hierarchical feature encoder and a conditionaldiscriminative network. Considering that the ultimate purpose is regressing objective score, we introduce a novel conditional cross entropy loss in the conditional-discriminative network to penalize the negative samples which hinder the convergence of the quality regression network. Experimental results show that the proposed method can achieve higher performance than traditional no-reference metrics, even comparable results with full-reference metrics. The proposed method also suggests the feasibility of assessing the quality of specific media content without the expensive and cumbersome subjective evaluations.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

No-Reference Point Cloud Quality Assessment via Domain Adaptation

Yang¹,

Liu²,

Chen³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Domain adaptation [40], [43], [44] is another potential way to alleviate the influence of distribution shift between synthetic and authentic distortions. In the IQA setting, Chen et al [3] developed the first Maximum Mean Discrepancy (MMD)-based domain adaptation for the quality assessment of screen content images (SCIs). However, different from ours, their methods were supervised and designed for human-rated data; the problem setting of ours is much more challenging.…”

Section: B Biqa In the Wildmentioning

confidence: 99%

“…Specifically, we first downloaded 750K images from Flicker 2 with diverse real-world distortions. The near-duplicate images were then removed by the command-line tool imgdupes() 3 , and those images of pure texts were also deleted, such as ppt, book pages, using OCR tool (pytesseract 4 ), and the images without Image Maker (e.g., Canon, Nikon, etc.) were de-duplicated.…”

Section: A Datasetsmentioning

confidence: 99%

Learning from Synthetic Data for Opinion-free Blind Image Quality Assessment in the Wild

Wang,

Tang,

Zhang

et al. 2021

Preprint

View full text Add to dashboard Cite

Nowadays, most existing blind image quality assessment (BIQA) models 1) are developed for synthetically-distorted images and often generalize poorly to authentic ones; 2) heavily rely on human ratings, which are prohibitively labor-expensive to collect. Here, we propose an opinion-f ree BIQA method that learns from synthetically-distorted images and multiple agents to assess the perceptual quality of authentically-distorted ones captured in the wild without relying on human labels. Specifically, we first assemble a large number of image pairs from syntheticallydistorted images and use a set of full-reference image quality assessment (FR-IQA) models to assign pseudo-binary labels of each pair indicating which image has higher quality as the supervisory signal. We then train a convolutional neural network (CNN)based BIQA model to rank the perceptual quality, optimized for consistency with the binary labels. Since there exists domain shift between the synthetically-and authentically-distorted images, an unsupervised domain adaptation (UDA) module is introduced to alleviate this issue. Extensive experiments demonstrate the effectiveness of our proposed opinion-f ree BIQA model, yielding state-of-the-art performance in terms of correlation with human opinion scores, as well as gMAD competition. Codes will be made publicly available upon acceptance.

show abstract

“…For BIQA whose annotation process is laborious and time-consuming, it's hard to obtain large amount of data, leading to the poor generalization under the cross-domain/crossdataset scenario. To alleviate the dependence of annotation, some works [1], [2] utilize the unsupervised domain adaptation (UDA) technique to transfer the learned knowledge from labelsufficient source domain to label-free target domain in IQA area. UCDA [1] adopts two-stage adaptation method, where in the first stage a coarse adaptation is applied between the source and target domain, and in the second stage a finelevel adaptation is applied between the confident and uncertain subdomains of target domain.…”

Section: Introductionmentioning

confidence: 99%

LIQA: Lifelong Blind Image Quality Assessment

Jian-zhao¹,

Zhou²,

Xu³

et al. 2023

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Existing learning-based methods for blind image quality assessment (BIQA) are heavily dependent on large amounts of annotated training data, and usually suffer from a severe performance degradation when encountering the domain/distribution shift problem. Thanks to the development of unsupervised domain adaptation (UDA), some works attempt to transfer the knowledge from a label-sufficient source domain to a label-free target domain under domain shift with UDA. However, it requires the coexistence of source and target data, which might be impractical for source data due to the privacy or storage issues. In this paper, we take the first step towards the source-free unsupervised domain adaptation (SFUDA) in a simple yet efficient manner for BIQA to tackle the domain shift without access to the source data. Specifically, we cast the quality assessment task as a rating distribution prediction problem. Based on the intrinsic properties of BIQA, we present a group of well-designed self-supervised objectives to guide the adaptation of the BN affine parameters towards the target domain. Among them, minimizing the prediction entropy and maximizing the batch prediction diversity aim to encourage more confident results while avoiding the trivial solution. Besides, based on the observation that the IQA rating distribution of single image follows the Gaussian distribution, we apply Gaussian regularization to the predicted rating distribution to make it more consistent with the nature of human scoring. Extensive experimental results under cross-domain scenarios demonstrated the effectiveness of our proposed method to mitigate the domain shift.

show abstract

No-Reference Screen Content Image Quality Assessment With Unsupervised Domain Adaptation

Cited by 28 publications

References 61 publications

No-Reference Point Cloud Quality Assessment via Domain Adaptation

No-Reference Point Cloud Quality Assessment via Domain Adaptation

Learning from Synthetic Data for Opinion-free Blind Image Quality Assessment in the Wild

LIQA: Lifelong Blind Image Quality Assessment

Contact Info

Product

Resources

About