MetaIQA: Deep Meta-Learning for No-Reference Image Quality Assessment

By Hancheng Zhu, Leida Li, Jinjian Wu, Weisheng Dong, Guangming Shi. Computer Vision and Pattern Recognition, 2020

386 citations recorded by the source corpus.

Recently, increasing interest has been drawn in exploiting deep convolutional neural networks (DCNNs) for no-reference image quality assessment (NR-IQA). Despite of the notable success achieved, there is a broad consensus that training DCNNs heavily relies on massive annotated data. Unfortunately, IQA is a typical small sample problem. Therefore, most of the existing DCNN-based IQA metrics operate based on pre-trained networks. However, these pre-trained networks are not designed for IQA task, leading to generalization problem when evaluating different types of distortions. With this motivation, this paper presents a no-reference IQA metric based on deep meta-learning. The underlying idea is to learn the meta-knowledge shared by human when evaluating the quality of images with various distortions, which can then be adapted to unknown distortions easily. Specifically, we first collect a number of NR-IQA tasks for different distortions. Then meta-learning is adopted to learn the prior knowledge shared by diversified distortions. Finally, the quality prior model is fine-tuned on a target NR-IQA task for quickly obtaining the quality model. Extensive experiments demonstrate that the proposed metric outperforms the state-of-the-arts by a large margin. Furthermore, the meta-model learned from synthetic distortions can also be easily generalized to authentic distortions, which is highly desired in real-world applications of IQA metrics.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/557e9371711c7409c78c96a6a2bea290a28cb365

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

No-Reference Image Quality Assessment via Transformers, Relative Ranking, and Self-Consistency (2021, background): “Zhu et al. [43] proposed a model to leverage meta-learning to learn the prior knowledge that is shared among different distortion types.”
Uncertainty-Aware Blind Image Quality Assessment in the Laboratory and Wild (2020, background): “[29] explored meta-learning as an advanced form of transfer learning in BIQA.”
TOPIQ: A Top-Down Approach From Semantics to Distortions for Image Quality Assessment (2023, methodology): “They are generally improved with more advanced network architecture, from deep belief net [36] to CNN [37], then to deeper CNN [38], [39], [40], later to ResNet [41], [42], [43], and now vision transformers [44], [45], [46].”
Blind Image Quality Assessment via Vision-Language Correspondence: A Multitask Learning Perspective (2023, background): “A different set of training tricks have been explored, including transfer learning [17, 76], meta learning [80], and contrastive learning [40].”
MUSIQ: Multi-scale Image Quality Transformer (2021, methodology): “We take numbers from [33, 48] for results of the reference methods.”
Exploring CLIP for Assessing the Look and Feel of Images (2022, methodology): “…from the rapid development of deep learning, many learning-based methods (Kang et al. 2014; Bosse et al. 2017; Ma et al. 2017b; Zhang et al. 2018b; Su et al. 2020; Zhu et al. 2020; Ke et al. 2021; Pan et al. 2022) have been proposed and shown superior performance compared with conventional methods.”
MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment (2022, methodology): “Pristine-quality images could be degraded in any stage of the whole circulation, from acquisition, compression, storage, transmission [61] to restoration by GAN-based algo-”