Matches in SemOpenAlex for { <https://semopenalex.org/work/W2998999750> ?p ?o ?g. }
- W2998999750 endingPage "39272" @default.
- W2998999750 startingPage "39254" @default.
- W2998999750 abstract "Image registration (or image alignment), the problem of aligning multiple images with relative displacement, is a crucial step in many multiframe image restoration algorithms. To solve the problem that most existing image registration approaches can only align two images in one inference, we propose a split-attention multiframe alignment network (SAMANet). Pixel-level displacements between multiple images are first estimated at low-resolution scales and then refined gradually with the increase in feature resolution. To better integrate the interframe information, we present a split-attention module (SAM) and a dot-product attention module (DPAM), which can adaptively rescale the cost volume features and optical flow features according to the similarity between features from different images. The experimental results demonstrate the superiority of our SAMANet over state-of-the-art image registration methods in terms of both accuracy and robustness. To solve the “ghosting effect” caused by pixelwise registration, we designed two “ghost” removal modules: warping repetition detection module (WRDM) and attention fusion module (AFM). WRDM detects “ghost” regions during the image warping process without increasing the time complexity of the registration algorithm. AFM uses an attention mechanism to rescale the aligned images and enables the registration network and the subsequent image restoration networks to be trained jointly. To validate the strengths of the proposed approaches, we apply SAMANet, WRDM and AFM to three image/video restoration tasks. Extensive evaluations demonstrate that the proposed methods can enhance the performance of image restoration algorithms and outperform the other compared registration algorithms." @default.
- W2998999750 created "2020-01-23" @default.
- W2998999750 creator A5022154076 @default.
- W2998999750 creator A5036398299 @default.
- W2998999750 creator A5052319916 @default.
- W2998999750 creator A5058687102 @default.
- W2998999750 creator A5070468133 @default.
- W2998999750 date "2020-01-01" @default.
- W2998999750 modified "2023-09-27" @default.
- W2998999750 title "Split-Attention Multiframe Alignment Network for Image Restoration" @default.
- W2998999750 cites W1513100184 @default.
- W2998999750 cites W1578285471 @default.
- W2998999750 cites W1677409904 @default.
- W2998999750 cites W1885185971 @default.
- W2998999750 cites W1951289974 @default.
- W2998999750 cites W1993120651 @default.
- W2998999750 cites W1993850091 @default.
- W2998999750 cites W2010981316 @default.
- W2998999750 cites W2031489346 @default.
- W2998999750 cites W2048695508 @default.
- W2998999750 cites W2056370875 @default.
- W2998999750 cites W2064076387 @default.
- W2998999750 cites W2080559363 @default.
- W2998999750 cites W2085261163 @default.
- W2998999750 cites W2111073598 @default.
- W2998999750 cites W2113221323 @default.
- W2998999750 cites W2114770744 @default.
- W2998999750 cites W2117228865 @default.
- W2998999750 cites W2119449517 @default.
- W2998999750 cites W2119473545 @default.
- W2998999750 cites W2143238378 @default.
- W2998999750 cites W2156956645 @default.
- W2998999750 cites W2242218935 @default.
- W2998999750 cites W2286655030 @default.
- W2998999750 cites W2320725294 @default.
- W2998999750 cites W2476548250 @default.
- W2998999750 cites W2478454054 @default.
- W2998999750 cites W2508457857 @default.
- W2998999750 cites W2557227117 @default.
- W2998999750 cites W2560474170 @default.
- W2998999750 cites W2575671312 @default.
- W2998999750 cites W2601564443 @default.
- W2998999750 cites W2610638008 @default.
- W2998999750 cites W2752782242 @default.
- W2998999750 cites W2770902190 @default.
- W2998999750 cites W2779333428 @default.
- W2998999750 cites W2798664922 @default.
- W2998999750 cites W2847088336 @default.
- W2998999750 cites W2866634454 @default.
- W2998999750 cites W2894877634 @default.
- W2998999750 cites W2894983388 @default.
- W2998999750 cites W2904340070 @default.
- W2998999750 cites W2908639863 @default.
- W2998999750 cites W2915130236 @default.
- W2998999750 cites W2920910272 @default.
- W2998999750 cites W2922435819 @default.
- W2998999750 cites W2938555542 @default.
- W2998999750 cites W2941051176 @default.
- W2998999750 cites W2952323569 @default.
- W2998999750 cites W2952825894 @default.
- W2998999750 cites W2954930822 @default.
- W2998999750 cites W2955058313 @default.
- W2998999750 cites W2962328014 @default.
- W2998999750 cites W2963091558 @default.
- W2998999750 cites W2963372104 @default.
- W2998999750 cites W2963548592 @default.
- W2998999750 cites W2963725279 @default.
- W2998999750 cites W2963782415 @default.
- W2998999750 cites W2963814095 @default.
- W2998999750 cites W2964040059 @default.
- W2998999750 cites W2964101377 @default.
- W2998999750 cites W2964125708 @default.
- W2998999750 cites W2964156315 @default.
- W2998999750 cites W2966926453 @default.
- W2998999750 cites W2967043539 @default.
- W2998999750 cites W3100388886 @default.
- W2998999750 cites W3102015846 @default.
- W2998999750 cites W3104725225 @default.
- W2998999750 cites W764651262 @default.
- W2998999750 doi "https://doi.org/10.1109/access.2020.2967028" @default.
- W2998999750 hasPublicationYear "2020" @default.
- W2998999750 type Work @default.
- W2998999750 sameAs 2998999750 @default.
- W2998999750 citedByCount "6" @default.
- W2998999750 countsByYear W29989997502021 @default.
- W2998999750 countsByYear W29989997502022 @default.
- W2998999750 countsByYear W29989997502023 @default.
- W2998999750 crossrefType "journal-article" @default.
- W2998999750 hasAuthorship W2998999750A5022154076 @default.
- W2998999750 hasAuthorship W2998999750A5036398299 @default.
- W2998999750 hasAuthorship W2998999750A5052319916 @default.
- W2998999750 hasAuthorship W2998999750A5058687102 @default.
- W2998999750 hasAuthorship W2998999750A5070468133 @default.
- W2998999750 hasBestOaLocation W29989997501 @default.
- W2998999750 hasConcept C104317684 @default.
- W2998999750 hasConcept C106430172 @default.
- W2998999750 hasConcept C115961682 @default.
- W2998999750 hasConcept C153180895 @default.