FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3037607532> ?p ?o ?g. }

• W3037607532 abstract "Deep kernel map networks have shown excellent performances in various classification problems including image annotation. Their general recipe consists in aggregating several layers of singular value decompositions (SVDs) -- that map data from input spaces into high dimensional spaces -- while preserving the similarity of the underlying kernels. However, the potential of these deep map networks has not been fully explored as the original setting of these networks focuses mainly on the approximation quality of their kernels and ignores their discrimination power. In this paper, we introduce a novel end-to-end design for deep kernel map learning that balances the approximation quality of kernels and their discrimination power. Our method proceeds in two steps; first, layerwise SVD is applied in order to build initial deep kernel map approximations and then an end-to-end supervised learning is employed to further enhance their discrimination power while maintaining their efficiency. Extensive experiments, conducted on the challenging ImageCLEF annotation benchmark, show the high efficiency and the out-performance of this two-step process with respect to different related methods." @default.
• W3037607532 created "2020-07-02" @default.
• W3037607532 creator A5017257056 @default.
• W3037607532 creator A5029896607 @default.
• W3037607532 date "2020-06-26" @default.
• W3037607532 modified "2023-09-27" @default.
• W3037607532 title "End-to-end training of deep kernel map networks for image classification" @default.
• W3037607532 cites W119301944 @default.
• W3037607532 cites W1510073064 @default.
• W3037607532 cites W1536680647 @default.
• W3037607532 cites W1540550673 @default.
• W3037607532 cites W1546411676 @default.
• W3037607532 cites W1563795667 @default.
• W3037607532 cites W1591394246 @default.
• W3037607532 cites W1591801644 @default.
• W3037607532 cites W1601437336 @default.
• W3037607532 cites W1604938182 @default.
• W3037607532 cites W1651008648 @default.
• W3037607532 cites W1698155719 @default.
• W3037607532 cites W1751437809 @default.
• W3037607532 cites W1869391892 @default.
• W3037607532 cites W1964357740 @default.
• W3037607532 cites W1965842729 @default.
• W3037607532 cites W1966347620 @default.
• W3037607532 cites W1967856816 @default.
• W3037607532 cites W1992208818 @default.
• W3037607532 cites W2017588182 @default.
• W3037607532 cites W2022137768 @default.
• W3037607532 cites W2022508996 @default.
• W3037607532 cites W2035976912 @default.
• W3037607532 cites W2039182213 @default.
• W3037607532 cites W2049033299 @default.
• W3037607532 cites W2067022618 @default.
• W3037607532 cites W2069797086 @default.
• W3037607532 cites W2070244236 @default.
• W3037607532 cites W2097117768 @default.
• W3037607532 cites W2102116870 @default.
• W3037607532 cites W2103914106 @default.
• W3037607532 cites W2105732805 @default.
• W3037607532 cites W2107968230 @default.
• W3037607532 cites W2109235804 @default.
• W3037607532 cites W2109798870 @default.
• W3037607532 cites W2112545207 @default.
• W3037607532 cites W2112562896 @default.
• W3037607532 cites W2112796928 @default.
• W3037607532 cites W2118585731 @default.
• W3037607532 cites W2123872146 @default.
• W3037607532 cites W2127069950 @default.
• W3037607532 cites W2131774270 @default.
• W3037607532 cites W2132285904 @default.
• W3037607532 cites W2136922672 @default.
• W3037607532 cites W2136940668 @default.
• W3037607532 cites W2137055149 @default.
• W3037607532 cites W2143338212 @default.
• W3037607532 cites W2144752499 @default.
• W3037607532 cites W2144902422 @default.
• W3037607532 cites W2146798037 @default.
• W3037607532 cites W2148492357 @default.
• W3037607532 cites W2148603752 @default.
• W3037607532 cites W2157801062 @default.
• W3037607532 cites W2163605009 @default.
• W3037607532 cites W2163922914 @default.
• W3037607532 cites W2167608136 @default.
• W3037607532 cites W2169278871 @default.
• W3037607532 cites W2169625877 @default.
• W3037607532 cites W2171937595 @default.
• W3037607532 cites W2172000360 @default.
• W3037607532 cites W2194775991 @default.
• W3037607532 cites W2222792052 @default.
• W3037607532 cites W2270330859 @default.
• W3037607532 cites W2279098554 @default.
• W3037607532 cites W2282078507 @default.
• W3037607532 cites W2294422256 @default.
• W3037607532 cites W2395459784 @default.
• W3037607532 cites W2399164823 @default.
• W3037607532 cites W2415785087 @default.
• W3037607532 cites W2485810183 @default.
• W3037607532 cites W2509715763 @default.
• W3037607532 cites W2562658572 @default.
• W3037607532 cites W2585201405 @default.
• W3037607532 cites W2587063199 @default.
• W3037607532 cites W2591766052 @default.
• W3037607532 cites W2595315035 @default.
• W3037607532 cites W2596743263 @default.
• W3037607532 cites W2618544249 @default.
• W3037607532 cites W2660819698 @default.
• W3037607532 cites W2755268668 @default.
• W3037607532 cites W2769171608 @default.
• W3037607532 cites W2770219845 @default.
• W3037607532 cites W2791891314 @default.
• W3037607532 cites W2793406320 @default.
• W3037607532 cites W2794055595 @default.
• W3037607532 cites W2892522582 @default.
• W3037607532 cites W2895163674 @default.
• W3037607532 cites W2903909270 @default.
• W3037607532 cites W2913150433 @default.
• W3037607532 cites W2919115771 @default.
• W3037607532 cites W2922027506 @default.
• W3037607532 cites W2940204692 @default.
• W3037607532 cites W2948436072 @default.

• next