FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313291231> ?p ?o ?g. }

• W4313291231 endingPage "693" @default.
• W4313291231 startingPage "682" @default.
• W4313291231 abstract "Weakly supervised object detection (WSOD) has received widespread attention since it requires only image-category annotations for detector training. Many advanced approaches solve this problem by a two-phase learning framework, that is, instance mining that classifies generated proposals via multiple instance learning, and instance refinement that iteratively refines bounding boxes using the supervision produced by the preceding stage. In this paper, we observe that the detection performance is usually limited by imprecise supervision, including part domination and untight boxes. To mitigate their adverse effects, we focus on selecting high-quality proposals as the supervision for WSOD. To be specific, for the issue of part domination, we propose bottom-up aggregated attention which incorporates low-level features from shallow layers to improve location representation of top-level features. In this manner, the proposals corresponding to entire objects can get high scores. Its advantage is that it can be flexibly plugged into the WSOD framework since there is no need to attach learnable parameters or learning branches. As regards the problem of untight boxes, we propose a phase-aware loss, which is the first work to measure supervision quality by the loss in the instance mining phase, to highlight correct boxes and suppress untight ones. In this work, we unify the proposed two modules into the framework of online instance classifier refinement. Extensive experiments on the PASCAL VOC and the MS COCO demonstrate that our method can significantly improve the performance of WSOD and achieve the state-of-the-art results. The code is available at <uri xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>https://github.com/Horatio9702/BUAA_PALoss</uri> ." @default.
• W4313291231 created "2023-01-06" @default.
• W4313291231 creator A5017617923 @default.
• W4313291231 creator A5026662451 @default.
• W4313291231 creator A5062318228 @default.
• W4313291231 creator A5063403522 @default.
• W4313291231 creator A5075541045 @default.
• W4313291231 creator A5083774648 @default.
• W4313291231 date "2023-01-01" @default.
• W4313291231 modified "2023-09-30" @default.
• W4313291231 title "Selecting High-Quality Proposals for Weakly Supervised Object Detection With Bottom-Up Aggregated Attention and Phase-Aware Loss" @default.
• W4313291231 cites W2031489346 @default.
• W4313291231 cites W2088049833 @default.
• W4313291231 cites W2108598243 @default.
• W4313291231 cites W2168804568 @default.
• W4313291231 cites W2194775991 @default.
• W4313291231 cites W2549139847 @default.
• W4313291231 cites W2565639579 @default.
• W4313291231 cites W2604260814 @default.
• W4313291231 cites W2752782242 @default.
• W4313291231 cites W2798748179 @default.
• W4313291231 cites W2813911573 @default.
• W4313291231 cites W2891894830 @default.
• W4313291231 cites W2895236117 @default.
• W4313291231 cites W2944938209 @default.
• W4313291231 cites W2954087924 @default.
• W4313291231 cites W2963150697 @default.
• W4313291231 cites W2963254348 @default.
• W4313291231 cites W2963351448 @default.
• W4313291231 cites W2963446712 @default.
• W4313291231 cites W2963516811 @default.
• W4313291231 cites W2963949812 @default.
• W4313291231 cites W2963952323 @default.
• W4313291231 cites W2964328846 @default.
• W4313291231 cites W2964444661 @default.
• W4313291231 cites W2969764577 @default.
• W4313291231 cites W2983943451 @default.
• W4313291231 cites W2987761193 @default.
• W4313291231 cites W2990400263 @default.
• W4313291231 cites W2991023920 @default.
• W4313291231 cites W2991533779 @default.
• W4313291231 cites W2991662170 @default.
• W4313291231 cites W2994041372 @default.
• W4313291231 cites W2997125355 @default.
• W4313291231 cites W3008354258 @default.
• W4313291231 cites W3017351200 @default.
• W4313291231 cites W3034329658 @default.
• W4313291231 cites W3034971973 @default.
• W4313291231 cites W3035725370 @default.
• W4313291231 cites W3093397049 @default.
• W4313291231 cites W3102701618 @default.
• W4313291231 cites W3117854388 @default.
• W4313291231 cites W3126494835 @default.
• W4313291231 cites W3175385356 @default.
• W4313291231 cites W3175704361 @default.
• W4313291231 cites W3190805089 @default.
• W4313291231 cites W3194159212 @default.
• W4313291231 cites W3194733604 @default.
• W4313291231 cites W3212216520 @default.
• W4313291231 cites W4224920452 @default.
• W4313291231 cites W4282913876 @default.
• W4313291231 cites W4312866875 @default.
• W4313291231 doi "https://doi.org/10.1109/tip.2022.3231744" @default.
• W4313291231 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37015622" @default.
• W4313291231 hasPublicationYear "2023" @default.
• W4313291231 type Work @default.
• W4313291231 citedByCount "6" @default.
• W4313291231 countsByYear W43132912312023 @default.
• W4313291231 crossrefType "journal-article" @default.
• W4313291231 hasAuthorship W4313291231A5017617923 @default.
• W4313291231 hasAuthorship W4313291231A5026662451 @default.
• W4313291231 hasAuthorship W4313291231A5062318228 @default.
• W4313291231 hasAuthorship W4313291231A5063403522 @default.
• W4313291231 hasAuthorship W4313291231A5075541045 @default.
• W4313291231 hasAuthorship W4313291231A5083774648 @default.
• W4313291231 hasConcept C111472728 @default.
• W4313291231 hasConcept C111919701 @default.
• W4313291231 hasConcept C119857082 @default.
• W4313291231 hasConcept C120665830 @default.
• W4313291231 hasConcept C121332964 @default.
• W4313291231 hasConcept C124101348 @default.
• W4313291231 hasConcept C136389625 @default.
• W4313291231 hasConcept C138885662 @default.
• W4313291231 hasConcept C153180895 @default.
• W4313291231 hasConcept C154945302 @default.
• W4313291231 hasConcept C17744445 @default.
• W4313291231 hasConcept C192209626 @default.
• W4313291231 hasConcept C199360897 @default.
• W4313291231 hasConcept C199539241 @default.
• W4313291231 hasConcept C2776151529 @default.
• W4313291231 hasConcept C2776359362 @default.
• W4313291231 hasConcept C2779530757 @default.
• W4313291231 hasConcept C41008148 @default.
• W4313291231 hasConcept C43126263 @default.
• W4313291231 hasConcept C50644808 @default.
• W4313291231 hasConcept C63584917 @default.
• W4313291231 hasConcept C75608658 @default.
• W4313291231 hasConcept C76155785 @default.

• next