Matches in SemOpenAlex for { <https://semopenalex.org/work/W4224852017> ?p ?o ?g. }
- W4224852017 abstract "Image quality assessment (IQA) algorithm aims to quantify the human perception of image quality. Unfortunately, there is a performance drop when assessing the distortion images generated by generative adversarial network (GAN) with seemingly realistic textures. In this work, we conjecture that this maladaptation lies in the backbone of IQA models, where patch-level prediction methods use independent image patches as input to calculate their scores separately, but lack spatial relationship modeling among image patches. Therefore, we propose an Attention-based Hybrid Image Quality Assessment Network (AHIQ) to deal with the challenge and get better performance on the GAN-based IQA task. Firstly, we adopt a two-branch architecture, including a vision transformer (ViT) branch and a convolutional neural network (CNN) branch for feature extraction. The hybrid architecture combines interaction information among image patches captured by ViT and local texture details from CNN. To make the features from the shallow CNN more focused on the visually salient region, a deformable convolution is applied with the help of semantic information from the ViT branch. Finally, we use a patch-wise score prediction module to obtain the final score. The experiments show that our model outperforms the state-of-the-art methods on four standard IQA datasets and AHIQ ranked first on the Full Reference (FR) track of the NTIRE 2022 Perceptual Image Quality Assessment Challenge. Code and pretrained models are publicly available at https://github.com/IIGROUP/AHIQ" @default.
- W4224852017 created "2022-04-27" @default.
- W4224852017 creator A5007566320 @default.
- W4224852017 creator A5020631593 @default.
- W4224852017 creator A5028709578 @default.
- W4224852017 creator A5063972231 @default.
- W4224852017 creator A5067285379 @default.
- W4224852017 creator A5085029347 @default.
- W4224852017 creator A5086329538 @default.
- W4224852017 creator A5086678836 @default.
- W4224852017 date "2022-06-01" @default.
- W4224852017 modified "2023-09-28" @default.
- W4224852017 title "Attentions Help CNNs See Better: Attention-based Hybrid Image Quality Assessment Network" @default.
- W4224852017 cites W1974013408 @default.
- W4224852017 cites W2015196405 @default.
- W4224852017 cites W2046119925 @default.
- W4224852017 cites W2067921844 @default.
- W4224852017 cites W2092264651 @default.
- W4224852017 cites W2117539524 @default.
- W4224852017 cites W2133665775 @default.
- W4224852017 cites W2144468361 @default.
- W4224852017 cites W2152059677 @default.
- W4224852017 cites W2161907179 @default.
- W4224852017 cites W2194775991 @default.
- W4224852017 cites W2319957123 @default.
- W4224852017 cites W2565312867 @default.
- W4224852017 cites W2571611310 @default.
- W4224852017 cites W2601564443 @default.
- W4224852017 cites W2737134362 @default.
- W4224852017 cites W2768814045 @default.
- W4224852017 cites W2798581339 @default.
- W4224852017 cites W2953590133 @default.
- W4224852017 cites W2962785568 @default.
- W4224852017 cites W2963598138 @default.
- W4224852017 cites W2964086649 @default.
- W4224852017 cites W2964350391 @default.
- W4224852017 cites W2965669158 @default.
- W4224852017 cites W2966926453 @default.
- W4224852017 cites W3009684847 @default.
- W4224852017 cites W3035022492 @default.
- W4224852017 cites W3035570181 @default.
- W4224852017 cites W3035719652 @default.
- W4224852017 cites W3036239693 @default.
- W4224852017 cites W3109016160 @default.
- W4224852017 cites W3174194560 @default.
- W4224852017 cites W3175517129 @default.
- W4224852017 cites W3176592254 @default.
- W4224852017 cites W3183401488 @default.
- W4224852017 cites W3194293177 @default.
- W4224852017 doi "https://doi.org/10.1109/cvprw56347.2022.00123" @default.
- W4224852017 hasPublicationYear "2022" @default.
- W4224852017 type Work @default.
- W4224852017 citedByCount "3" @default.
- W4224852017 countsByYear W42248520172023 @default.
- W4224852017 crossrefType "proceedings-article" @default.
- W4224852017 hasAuthorship W4224852017A5007566320 @default.
- W4224852017 hasAuthorship W4224852017A5020631593 @default.
- W4224852017 hasAuthorship W4224852017A5028709578 @default.
- W4224852017 hasAuthorship W4224852017A5063972231 @default.
- W4224852017 hasAuthorship W4224852017A5067285379 @default.
- W4224852017 hasAuthorship W4224852017A5085029347 @default.
- W4224852017 hasAuthorship W4224852017A5086329538 @default.
- W4224852017 hasAuthorship W4224852017A5086678836 @default.
- W4224852017 hasBestOaLocation W42248520172 @default.
- W4224852017 hasConcept C108583219 @default.
- W4224852017 hasConcept C115961682 @default.
- W4224852017 hasConcept C138885662 @default.
- W4224852017 hasConcept C153180895 @default.
- W4224852017 hasConcept C154945302 @default.
- W4224852017 hasConcept C2776401178 @default.
- W4224852017 hasConcept C31972630 @default.
- W4224852017 hasConcept C41008148 @default.
- W4224852017 hasConcept C41895202 @default.
- W4224852017 hasConcept C52622490 @default.
- W4224852017 hasConcept C55020928 @default.
- W4224852017 hasConcept C81363708 @default.
- W4224852017 hasConceptScore W4224852017C108583219 @default.
- W4224852017 hasConceptScore W4224852017C115961682 @default.
- W4224852017 hasConceptScore W4224852017C138885662 @default.
- W4224852017 hasConceptScore W4224852017C153180895 @default.
- W4224852017 hasConceptScore W4224852017C154945302 @default.
- W4224852017 hasConceptScore W4224852017C2776401178 @default.
- W4224852017 hasConceptScore W4224852017C31972630 @default.
- W4224852017 hasConceptScore W4224852017C41008148 @default.
- W4224852017 hasConceptScore W4224852017C41895202 @default.
- W4224852017 hasConceptScore W4224852017C52622490 @default.
- W4224852017 hasConceptScore W4224852017C55020928 @default.
- W4224852017 hasConceptScore W4224852017C81363708 @default.
- W4224852017 hasFunder F4320321001 @default.
- W4224852017 hasLocation W42248520171 @default.
- W4224852017 hasLocation W42248520172 @default.
- W4224852017 hasOpenAccess W4224852017 @default.
- W4224852017 hasPrimaryLocation W42248520171 @default.
- W4224852017 hasRelatedWork W2279398222 @default.
- W4224852017 hasRelatedWork W2546942002 @default.
- W4224852017 hasRelatedWork W2731899572 @default.
- W4224852017 hasRelatedWork W3133861977 @default.
- W4224852017 hasRelatedWork W3156786002 @default.
- W4224852017 hasRelatedWork W4200173597 @default.
- W4224852017 hasRelatedWork W4299822940 @default.