Matches in SemOpenAlex for { <https://semopenalex.org/work/W4293784368> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4293784368 abstract "This paper aims at exploring how to synthesize close-to-real blurs that existing video deblurring models trained on them can generalize well to real-world blurry videos. In recent years, deep learning-based approaches have achieved promising success on video deblurring task. However, the models trained on existing synthetic datasets still suffer from generalization problems over real-world blurry scenarios with undesired artifacts. The factors accounting for the failure remain unknown. Therefore, we revisit the classical blur synthesis pipeline and figure out the possible reasons, including shooting parameters, blur formation space, and image signal processor~(ISP). To analyze the effects of these potential factors, we first collect an ultra-high frame-rate (940 FPS) RAW video dataset as the data basis to synthesize various kinds of blurs. Then we propose a novel realistic blur synthesis pipeline termed as RAW-Blur by leveraging blur formation cues. Through numerous experiments, we demonstrate that synthesizing blurs in the RAW space and adopting the same ISP as the real-world testing data can effectively eliminate the negative effects of synthetic data. Furthermore, the shooting parameters of the synthesized blurry video, e.g., exposure time and frame-rate play significant roles in improving the performance of deblurring models. Impressively, the models trained on the blurry data synthesized by the proposed RAW-Blur pipeline can obtain more than 5dB PSNR gain against those trained on the existing synthetic blur datasets. We believe the novel realistic synthesis pipeline and the corresponding RAW video dataset can help the community to easily construct customized blur datasets to improve real-world video deblurring performance largely, instead of laboriously collecting real data pairs." @default.
- W4293784368 created "2022-08-31" @default.
- W4293784368 creator A5007566320 @default.
- W4293784368 creator A5013983776 @default.
- W4293784368 creator A5020953714 @default.
- W4293784368 creator A5048033343 @default.
- W4293784368 creator A5063041433 @default.
- W4293784368 creator A5064222660 @default.
- W4293784368 creator A5076036593 @default.
- W4293784368 creator A5087622795 @default.
- W4293784368 date "2022-08-28" @default.
- W4293784368 modified "2023-10-15" @default.
- W4293784368 title "Towards Real-World Video Deblurring by Exploring Blur Formation Process" @default.
- W4293784368 doi "https://doi.org/10.48550/arxiv.2208.13184" @default.
- W4293784368 hasPublicationYear "2022" @default.
- W4293784368 type Work @default.
- W4293784368 citedByCount "0" @default.
- W4293784368 crossrefType "posted-content" @default.
- W4293784368 hasAuthorship W4293784368A5007566320 @default.
- W4293784368 hasAuthorship W4293784368A5013983776 @default.
- W4293784368 hasAuthorship W4293784368A5020953714 @default.
- W4293784368 hasAuthorship W4293784368A5048033343 @default.
- W4293784368 hasAuthorship W4293784368A5063041433 @default.
- W4293784368 hasAuthorship W4293784368A5064222660 @default.
- W4293784368 hasAuthorship W4293784368A5076036593 @default.
- W4293784368 hasAuthorship W4293784368A5087622795 @default.
- W4293784368 hasBestOaLocation W42937843681 @default.
- W4293784368 hasConcept C106430172 @default.
- W4293784368 hasConcept C111919701 @default.
- W4293784368 hasConcept C115961682 @default.
- W4293784368 hasConcept C126042441 @default.
- W4293784368 hasConcept C134306372 @default.
- W4293784368 hasConcept C154945302 @default.
- W4293784368 hasConcept C177148314 @default.
- W4293784368 hasConcept C199360897 @default.
- W4293784368 hasConcept C2777693668 @default.
- W4293784368 hasConcept C2777708103 @default.
- W4293784368 hasConcept C31972630 @default.
- W4293784368 hasConcept C3261483 @default.
- W4293784368 hasConcept C33923547 @default.
- W4293784368 hasConcept C41008148 @default.
- W4293784368 hasConcept C43521106 @default.
- W4293784368 hasConcept C76155785 @default.
- W4293784368 hasConcept C9417928 @default.
- W4293784368 hasConcept C98045186 @default.
- W4293784368 hasConceptScore W4293784368C106430172 @default.
- W4293784368 hasConceptScore W4293784368C111919701 @default.
- W4293784368 hasConceptScore W4293784368C115961682 @default.
- W4293784368 hasConceptScore W4293784368C126042441 @default.
- W4293784368 hasConceptScore W4293784368C134306372 @default.
- W4293784368 hasConceptScore W4293784368C154945302 @default.
- W4293784368 hasConceptScore W4293784368C177148314 @default.
- W4293784368 hasConceptScore W4293784368C199360897 @default.
- W4293784368 hasConceptScore W4293784368C2777693668 @default.
- W4293784368 hasConceptScore W4293784368C2777708103 @default.
- W4293784368 hasConceptScore W4293784368C31972630 @default.
- W4293784368 hasConceptScore W4293784368C3261483 @default.
- W4293784368 hasConceptScore W4293784368C33923547 @default.
- W4293784368 hasConceptScore W4293784368C41008148 @default.
- W4293784368 hasConceptScore W4293784368C43521106 @default.
- W4293784368 hasConceptScore W4293784368C76155785 @default.
- W4293784368 hasConceptScore W4293784368C9417928 @default.
- W4293784368 hasConceptScore W4293784368C98045186 @default.
- W4293784368 hasLocation W42937843681 @default.
- W4293784368 hasOpenAccess W4293784368 @default.
- W4293784368 hasPrimaryLocation W42937843681 @default.
- W4293784368 hasRelatedWork W2020789691 @default.
- W4293784368 hasRelatedWork W2050607296 @default.
- W4293784368 hasRelatedWork W2060613391 @default.
- W4293784368 hasRelatedWork W2080322084 @default.
- W4293784368 hasRelatedWork W2120286341 @default.
- W4293784368 hasRelatedWork W2144553684 @default.
- W4293784368 hasRelatedWork W2354226224 @default.
- W4293784368 hasRelatedWork W2579111433 @default.
- W4293784368 hasRelatedWork W2738579427 @default.
- W4293784368 hasRelatedWork W3119379899 @default.
- W4293784368 isParatext "false" @default.
- W4293784368 isRetracted "false" @default.
- W4293784368 workType "article" @default.