FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3029299740> ?p ?o ?g. }

• W3029299740 abstract "Existing CNN-based methods for pixel labeling heavily depend on multi-scale features to meet the requirements of both semantic comprehension and detail preservation. State-of-the-art pixel labeling neural networks widely exploit conventional scale-transfer operations, i.e., up-sampling and down-sampling to learn multi-scale features. In this work, we find that these operations lead to scale-confused features and suboptimal performance because they are spatial-invariant and directly transit all feature information cross scales without spatial selection. To address this issue, we propose the Gated Scale-Transfer Operation (GSTO) to properly transit spatial-filtered features to another scale. Specifically, GSTO can work either with or without extra supervision. Unsupervised GSTO is learned from the feature itself while the supervised one is guided by the supervised probability matrix. Both forms of GSTO are lightweight and plug-and-play, which can be flexibly integrated into networks or modules for learning better multi-scale features. In particular, by plugging GSTO into HRNet, we get a more powerful backbone (namely GSTO-HRNet) for pixel labeling, and it achieves new state-of-the-art results on the COCO benchmark for human pose estimation and other benchmarks for semantic segmentation including Cityscapes, LIP and Pascal Context, with negligible extra computational cost. Moreover, experiment results demonstrate that GSTO can also significantly boost the performance of multi-scale feature aggregation modules like PPM and ASPP. Code will be made available at https://github.com/VDIGPKU/GSTO." @default.
• W3029299740 created "2020-06-05" @default.
• W3029299740 creator A5020650971 @default.
• W3029299740 creator A5031953394 @default.
• W3029299740 creator A5052945216 @default.
• W3029299740 creator A5053275663 @default.
• W3029299740 creator A5061469520 @default.
• W3029299740 creator A5074449757 @default.
• W3029299740 creator A5089296571 @default.
• W3029299740 date "2020-05-27" @default.
• W3029299740 modified "2023-09-26" @default.
• W3029299740 title "GSTO: Gated Scale-Transfer Operation for Multi-Scale Feature Learning in Pixel Labeling" @default.
• W3029299740 cites W1495267108 @default.
• W3029299740 cites W1861492603 @default.
• W3029299740 cites W1903029394 @default.
• W3029299740 cites W2125215748 @default.
• W3029299740 cites W2307770531 @default.
• W3029299740 cites W2340897893 @default.
• W3029299740 cites W2412782625 @default.
• W3029299740 cites W2476548250 @default.
• W3029299740 cites W2555751471 @default.
• W3029299740 cites W2557406251 @default.
• W3029299740 cites W2559085405 @default.
• W3029299740 cites W2560023338 @default.
• W3029299740 cites W2563705555 @default.
• W3029299740 cites W2578797046 @default.
• W3029299740 cites W2598666589 @default.
• W3029299740 cites W2598915960 @default.
• W3029299740 cites W2630837129 @default.
• W3029299740 cites W2752782242 @default.
• W3029299740 cites W2798791840 @default.
• W3029299740 cites W2799166040 @default.
• W3029299740 cites W2799213142 @default.
• W3029299740 cites W2804488433 @default.
• W3029299740 cites W2819476901 @default.
• W3029299740 cites W2883779225 @default.
• W3029299740 cites W2884436604 @default.
• W3029299740 cites W2884585870 @default.
• W3029299740 cites W2886934227 @default.
• W3029299740 cites W2888340395 @default.
• W3029299740 cites W2890319229 @default.
• W3029299740 cites W2894913190 @default.
• W3029299740 cites W2916798096 @default.
• W3029299740 cites W2921781974 @default.
• W3029299740 cites W2940262938 @default.
• W3029299740 cites W2962773068 @default.
• W3029299740 cites W2962802951 @default.
• W3029299740 cites W2962891704 @default.
• W3029299740 cites W2963150697 @default.
• W3029299740 cites W2963402313 @default.
• W3029299740 cites W2963495494 @default.
• W3029299740 cites W2963598138 @default.
• W3029299740 cites W2963727650 @default.
• W3029299740 cites W2963781481 @default.
• W3029299740 cites W2963843116 @default.
• W3029299740 cites W2963978393 @default.
• W3029299740 cites W2964221239 @default.
• W3029299740 cites W2964252655 @default.
• W3029299740 cites W2964288706 @default.
• W3029299740 cites W2964309882 @default.
• W3029299740 cites W2990032492 @default.
• W3029299740 cites W2990775046 @default.
• W3029299740 cites W2991471181 @default.
• W3029299740 doi "https://doi.org/10.48550/arxiv.2005.13363" @default.
• W3029299740 hasPublicationYear "2020" @default.
• W3029299740 type Work @default.
• W3029299740 sameAs 3029299740 @default.
• W3029299740 citedByCount "0" @default.
• W3029299740 crossrefType "posted-content" @default.
• W3029299740 hasAuthorship W3029299740A5020650971 @default.
• W3029299740 hasAuthorship W3029299740A5031953394 @default.
• W3029299740 hasAuthorship W3029299740A5052945216 @default.
• W3029299740 hasAuthorship W3029299740A5053275663 @default.
• W3029299740 hasAuthorship W3029299740A5061469520 @default.
• W3029299740 hasAuthorship W3029299740A5074449757 @default.
• W3029299740 hasAuthorship W3029299740A5089296571 @default.
• W3029299740 hasBestOaLocation W30292997401 @default.
• W3029299740 hasConcept C119857082 @default.
• W3029299740 hasConcept C121332964 @default.
• W3029299740 hasConcept C13280743 @default.
• W3029299740 hasConcept C138885662 @default.
• W3029299740 hasConcept C151730666 @default.
• W3029299740 hasConcept C153180895 @default.
• W3029299740 hasConcept C154945302 @default.
• W3029299740 hasConcept C160633673 @default.
• W3029299740 hasConcept C165696696 @default.
• W3029299740 hasConcept C185798385 @default.
• W3029299740 hasConcept C199360897 @default.
• W3029299740 hasConcept C205649164 @default.
• W3029299740 hasConcept C2776401178 @default.
• W3029299740 hasConcept C2778755073 @default.
• W3029299740 hasConcept C2779343474 @default.
• W3029299740 hasConcept C38652104 @default.
• W3029299740 hasConcept C41008148 @default.
• W3029299740 hasConcept C41895202 @default.
• W3029299740 hasConcept C59404180 @default.
• W3029299740 hasConcept C62520636 @default.
• W3029299740 hasConcept C75608658 @default.
• W3029299740 hasConcept C86803240 @default.
• W3029299740 hasConcept C89600930 @default.

• next