FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3149775056> ?p ?o ?g. }

• W3149775056 endingPage "1" @default.
• W3149775056 startingPage "1" @default.
• W3149775056 abstract "The research in image quality assessment (IQA) has a long history, and significant progress has been made by leveraging recent advances in deep neural networks (DNNs). Despite high correlation numbers on existing IQA datasets, DNN-based models may be easily falsified in the group maximum differentiation (gMAD) competition. Here we show that gMAD examples can be used to improve blind IQA (BIQA) methods. Specifically, we first pre-train a DNN-based BIQA model using multiple noisy annotators, and fine-tune it on multiple synthetically distorted images, resulting in a top-performing baseline model. We then seek pairs of images by comparing the baseline model with a set of full-reference IQA methods in gMAD. The spotted gMAD examples are most likely to reveal the weaknesses of the baseline, and suggest potential ways for refinement. We query human quality annotations for the selected images in a well-controlled laboratory environment, and further fine-tune the baseline on the combination of human-rated images from gMAD and existing databases. This process may be iterated, enabling active fine-tuning from gMAD examples for BIQA. We demonstrate the feasibility of our active learning scheme on a large-scale unlabeled image set, and show that the fine-tuned quality model achieves improved generalizability in gMAD, without destroying performance on previously seen databases." @default.
• W3149775056 created "2021-04-13" @default.
• W3149775056 creator A5020029652 @default.
• W3149775056 creator A5024076883 @default.
• W3149775056 date "2021-01-01" @default.
• W3149775056 modified "2023-10-12" @default.
• W3149775056 title "Active Fine-Tuning from gMAD Examples Improves Blind Image Quality Assessment" @default.
• W3149775056 cites W109907426 @default.
• W3149775056 cites W1580389772 @default.
• W3149775056 cites W1974013408 @default.
• W3149775056 cites W1982471090 @default.
• W3149775056 cites W1990873664 @default.
• W3149775056 cites W2013422999 @default.
• W3149775056 cites W2015196405 @default.
• W3149775056 cites W2031614119 @default.
• W3149775056 cites W2033442452 @default.
• W3149775056 cites W2046119925 @default.
• W3149775056 cites W2053005719 @default.
• W3149775056 cites W2078807908 @default.
• W3149775056 cites W2085518012 @default.
• W3149775056 cites W2102166818 @default.
• W3149775056 cites W2114338738 @default.
• W3149775056 cites W2126226185 @default.
• W3149775056 cites W2129644086 @default.
• W3149775056 cites W2133665775 @default.
• W3149775056 cites W2138790992 @default.
• W3149775056 cites W2151035455 @default.
• W3149775056 cites W2161907179 @default.
• W3149775056 cites W2162915697 @default.
• W3149775056 cites W2168078737 @default.
• W3149775056 cites W2168356304 @default.
• W3149775056 cites W2170319235 @default.
• W3149775056 cites W2294857031 @default.
• W3149775056 cites W2473697052 @default.
• W3149775056 cites W2513500606 @default.
• W3149775056 cites W2556068545 @default.
• W3149775056 cites W2565312867 @default.
• W3149775056 cites W2571611310 @default.
• W3149775056 cites W2768340063 @default.
• W3149775056 cites W2770961662 @default.
• W3149775056 cites W2798581339 @default.
• W3149775056 cites W2807793257 @default.
• W3149775056 cites W2903193781 @default.
• W3149775056 cites W2905544033 @default.
• W3149775056 cites W2906729185 @default.
• W3149775056 cites W2953590133 @default.
• W3149775056 cites W2963975576 @default.
• W3149775056 cites W2970763616 @default.
• W3149775056 cites W3035595647 @default.
• W3149775056 cites W3090767700 @default.
• W3149775056 cites W3091249416 @default.
• W3149775056 cites W3098560717 @default.
• W3149775056 cites W3100498948 @default.
• W3149775056 cites W3102733987 @default.
• W3149775056 cites W3103635814 @default.
• W3149775056 doi "https://doi.org/10.1109/tpami.2021.3071759" @default.
• W3149775056 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/33830918" @default.
• W3149775056 hasPublicationYear "2021" @default.
• W3149775056 type Work @default.
• W3149775056 sameAs 3149775056 @default.
• W3149775056 citedByCount "8" @default.
• W3149775056 countsByYear W31497750562021 @default.
• W3149775056 countsByYear W31497750562022 @default.
• W3149775056 countsByYear W31497750562023 @default.
• W3149775056 crossrefType "journal-article" @default.
• W3149775056 hasAuthorship W3149775056A5020029652 @default.
• W3149775056 hasAuthorship W3149775056A5024076883 @default.
• W3149775056 hasBestOaLocation W31497750562 @default.
• W3149775056 hasConcept C105795698 @default.
• W3149775056 hasConcept C111368507 @default.
• W3149775056 hasConcept C111472728 @default.
• W3149775056 hasConcept C111919701 @default.
• W3149775056 hasConcept C115961682 @default.
• W3149775056 hasConcept C119857082 @default.
• W3149775056 hasConcept C12725497 @default.
• W3149775056 hasConcept C127313418 @default.
• W3149775056 hasConcept C138885662 @default.
• W3149775056 hasConcept C153180895 @default.
• W3149775056 hasConcept C154945302 @default.
• W3149775056 hasConcept C162324750 @default.
• W3149775056 hasConcept C176217482 @default.
• W3149775056 hasConcept C177264268 @default.
• W3149775056 hasConcept C199360897 @default.
• W3149775056 hasConcept C21547014 @default.
• W3149775056 hasConcept C27158222 @default.
• W3149775056 hasConcept C2779346075 @default.
• W3149775056 hasConcept C2779530757 @default.
• W3149775056 hasConcept C33923547 @default.
• W3149775056 hasConcept C41008148 @default.
• W3149775056 hasConcept C55020928 @default.
• W3149775056 hasConcept C98045186 @default.
• W3149775056 hasConceptScore W3149775056C105795698 @default.
• W3149775056 hasConceptScore W3149775056C111368507 @default.
• W3149775056 hasConceptScore W3149775056C111472728 @default.
• W3149775056 hasConceptScore W3149775056C111919701 @default.
• W3149775056 hasConceptScore W3149775056C115961682 @default.
• W3149775056 hasConceptScore W3149775056C119857082 @default.
• W3149775056 hasConceptScore W3149775056C12725497 @default.

• next