Matches in SemOpenAlex for { <https://semopenalex.org/work/W3035682321> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W3035682321 abstract "Neural architecture search (NAS) advances beyond the state-of-the-art in various computer vision tasks by automating the designs of deep neural networks. In this paper, we aim to address three important questions in NAS: (1) How to measure the correlation between architectures and their performances? (2) How to evaluate the correlation between different architectures? (3) How to learn these correlations with a small number of samples? To this end, we first model these correlations from a Bayesian perspective. Specifically, by introducing a novel Gaussian Process based NAS (GP-NAS) method, the correlations are modeled by the kernel function and mean function. The kernel function is also learnable to enable adaptive modeling for complex correlations in different search spaces. Furthermore, by incorporating a mutual information based sampling method, we can theoretically ensure the high-performance architecture with only a small set of samples. After addressing these problems, training GP-NAS once enables direct performance prediction of any architecture in different scenarios and may obtain efficient networks for different deployment platforms. Extensive experiments on both image classification and face recognition tasks verify the effectiveness of our algorithm." @default.
- W3035682321 created "2020-06-19" @default.
- W3035682321 creator A5028517540 @default.
- W3035682321 creator A5032038866 @default.
- W3035682321 creator A5034194826 @default.
- W3035682321 creator A5059407466 @default.
- W3035682321 creator A5073110676 @default.
- W3035682321 creator A5075717206 @default.
- W3035682321 date "2020-06-01" @default.
- W3035682321 modified "2023-10-01" @default.
- W3035682321 title "GP-NAS: Gaussian Process Based Neural Architecture Search" @default.
- W3035682321 cites W2019827586 @default.
- W3035682321 cites W2117539524 @default.
- W3035682321 cites W2796265726 @default.
- W3035682321 cites W2921495890 @default.
- W3035682321 cites W2955051405 @default.
- W3035682321 cites W2960010704 @default.
- W3035682321 cites W2961666066 @default.
- W3035682321 cites W2962919941 @default.
- W3035682321 cites W2963136578 @default.
- W3035682321 cites W2963163009 @default.
- W3035682321 cites W2964081807 @default.
- W3035682321 cites W2964444661 @default.
- W3035682321 cites W2965658867 @default.
- W3035682321 cites W2967733054 @default.
- W3035682321 cites W2969985801 @default.
- W3035682321 cites W2980270353 @default.
- W3035682321 cites W2981383995 @default.
- W3035682321 cites W2981748264 @default.
- W3035682321 cites W2982083293 @default.
- W3035682321 cites W2988089665 @default.
- W3035682321 cites W4232464081 @default.
- W3035682321 cites W4238682518 @default.
- W3035682321 doi "https://doi.org/10.1109/cvpr42600.2020.01195" @default.
- W3035682321 hasPublicationYear "2020" @default.
- W3035682321 type Work @default.
- W3035682321 sameAs 3035682321 @default.
- W3035682321 citedByCount "33" @default.
- W3035682321 countsByYear W30356823212020 @default.
- W3035682321 countsByYear W30356823212021 @default.
- W3035682321 countsByYear W30356823212022 @default.
- W3035682321 countsByYear W30356823212023 @default.
- W3035682321 crossrefType "proceedings-article" @default.
- W3035682321 hasAuthorship W3035682321A5028517540 @default.
- W3035682321 hasAuthorship W3035682321A5032038866 @default.
- W3035682321 hasAuthorship W3035682321A5034194826 @default.
- W3035682321 hasAuthorship W3035682321A5059407466 @default.
- W3035682321 hasAuthorship W3035682321A5073110676 @default.
- W3035682321 hasAuthorship W3035682321A5075717206 @default.
- W3035682321 hasConcept C119857082 @default.
- W3035682321 hasConcept C121332964 @default.
- W3035682321 hasConcept C123657996 @default.
- W3035682321 hasConcept C142362112 @default.
- W3035682321 hasConcept C153180895 @default.
- W3035682321 hasConcept C153349607 @default.
- W3035682321 hasConcept C154945302 @default.
- W3035682321 hasConcept C163716315 @default.
- W3035682321 hasConcept C199360897 @default.
- W3035682321 hasConcept C41008148 @default.
- W3035682321 hasConcept C61326573 @default.
- W3035682321 hasConcept C62520636 @default.
- W3035682321 hasConcept C98045186 @default.
- W3035682321 hasConceptScore W3035682321C119857082 @default.
- W3035682321 hasConceptScore W3035682321C121332964 @default.
- W3035682321 hasConceptScore W3035682321C123657996 @default.
- W3035682321 hasConceptScore W3035682321C142362112 @default.
- W3035682321 hasConceptScore W3035682321C153180895 @default.
- W3035682321 hasConceptScore W3035682321C153349607 @default.
- W3035682321 hasConceptScore W3035682321C154945302 @default.
- W3035682321 hasConceptScore W3035682321C163716315 @default.
- W3035682321 hasConceptScore W3035682321C199360897 @default.
- W3035682321 hasConceptScore W3035682321C41008148 @default.
- W3035682321 hasConceptScore W3035682321C61326573 @default.
- W3035682321 hasConceptScore W3035682321C62520636 @default.
- W3035682321 hasConceptScore W3035682321C98045186 @default.
- W3035682321 hasLocation W30356823211 @default.
- W3035682321 hasOpenAccess W3035682321 @default.
- W3035682321 hasPrimaryLocation W30356823211 @default.
- W3035682321 hasRelatedWork W2130075347 @default.
- W3035682321 hasRelatedWork W2364531466 @default.
- W3035682321 hasRelatedWork W2783038087 @default.
- W3035682321 hasRelatedWork W2961085424 @default.
- W3035682321 hasRelatedWork W3046775127 @default.
- W3035682321 hasRelatedWork W3170094116 @default.
- W3035682321 hasRelatedWork W3209574120 @default.
- W3035682321 hasRelatedWork W4205958290 @default.
- W3035682321 hasRelatedWork W4286629047 @default.
- W3035682321 hasRelatedWork W4224009465 @default.
- W3035682321 isParatext "false" @default.
- W3035682321 isRetracted "false" @default.
- W3035682321 magId "3035682321" @default.
- W3035682321 workType "article" @default.