Matches in SemOpenAlex for { <https://semopenalex.org/work/W2892058377> ?p ?o ?g. }
Showing items 1 to 61 of
61
with 100 items per page.
- W2892058377 abstract "Deep learning is a rapidly growing discipline that models high-level features in data as multilayeredneural networks. The recent trend toward deep neural networks has been driven, in large part, bya combination of affordable computing hardware, open source software, and the availability ofpre-trained networks on large-scale datasets.In this thesis, we propose deep learning approaches to 3D shape recognition using a multilevelfeature learning paradigm. We start by comprehensively reviewing recent shape descriptors,including hand-crafted descriptors that are mostly developed in the spectral geometry setting andalso the ones obtained via learning-based methods. Then, we introduce novel multi-level featurelearning approaches using spectral graph wavelets, bag-of-features and deep learning. Low-levelfeatures are first extracted from a 3D shape using spectral graph wavelets. Mid-level features arethen generated via the bag-of-features model by employing locality-constrained linear coding as afeature coding method, in conjunction with the biharmonic distance and intrinsic spatial pyramidmatching in a bid to effectively measure the spatial relationship between each pair of the bag-offeaturedescriptors.For the task of 3D shape retrieval, high-level shape features are learned via a deep auto-encoderon mid-level features. Then, we compare the deep learned descriptor of a query shape to thedescriptors of all shapes in the dataset using a dissimilarity measure for 3D shape retrieval. For thetask of 3D shape classification, mid-level features are represented as 2D images in order to be fedinto a pre-trained convolutional neural network to learn high-level features from the penultimatefully-connected layer of the network. Finally, a multiclass support vector machine classifier istrained on these deep learned descriptors, and the classification accuracy is subsequently computed.The proposed 3D shape retrieval and classification approaches are evaluated on three standard 3Dshape benchmarks through extensive experiments, and the results show compelling superiority ofour approaches over state-of-the-art methods." @default.
- W2892058377 created "2018-09-27" @default.
- W2892058377 creator A5068826409 @default.
- W2892058377 date "2017-10-23" @default.
- W2892058377 modified "2023-09-27" @default.
- W2892058377 title "Deep Shape Representations for 3D Object Recognition" @default.
- W2892058377 hasPublicationYear "2017" @default.
- W2892058377 type Work @default.
- W2892058377 sameAs 2892058377 @default.
- W2892058377 citedByCount "0" @default.
- W2892058377 crossrefType "dissertation" @default.
- W2892058377 hasAuthorship W2892058377A5068826409 @default.
- W2892058377 hasConcept C108583219 @default.
- W2892058377 hasConcept C121684516 @default.
- W2892058377 hasConcept C138885662 @default.
- W2892058377 hasConcept C153180895 @default.
- W2892058377 hasConcept C154945302 @default.
- W2892058377 hasConcept C2776401178 @default.
- W2892058377 hasConcept C31487907 @default.
- W2892058377 hasConcept C41008148 @default.
- W2892058377 hasConcept C41895202 @default.
- W2892058377 hasConcept C50644808 @default.
- W2892058377 hasConcept C59404180 @default.
- W2892058377 hasConceptScore W2892058377C108583219 @default.
- W2892058377 hasConceptScore W2892058377C121684516 @default.
- W2892058377 hasConceptScore W2892058377C138885662 @default.
- W2892058377 hasConceptScore W2892058377C153180895 @default.
- W2892058377 hasConceptScore W2892058377C154945302 @default.
- W2892058377 hasConceptScore W2892058377C2776401178 @default.
- W2892058377 hasConceptScore W2892058377C31487907 @default.
- W2892058377 hasConceptScore W2892058377C41008148 @default.
- W2892058377 hasConceptScore W2892058377C41895202 @default.
- W2892058377 hasConceptScore W2892058377C50644808 @default.
- W2892058377 hasConceptScore W2892058377C59404180 @default.
- W2892058377 hasLocation W28920583771 @default.
- W2892058377 hasOpenAccess W2892058377 @default.
- W2892058377 hasPrimaryLocation W28920583771 @default.
- W2892058377 hasRelatedWork W1958841061 @default.
- W2892058377 hasRelatedWork W2074880021 @default.
- W2892058377 hasRelatedWork W2279608790 @default.
- W2892058377 hasRelatedWork W2300169576 @default.
- W2892058377 hasRelatedWork W2785412887 @default.
- W2892058377 hasRelatedWork W2807092842 @default.
- W2892058377 hasRelatedWork W2887082564 @default.
- W2892058377 hasRelatedWork W2889589660 @default.
- W2892058377 hasRelatedWork W2902344891 @default.
- W2892058377 hasRelatedWork W2908240258 @default.
- W2892058377 hasRelatedWork W2995795169 @default.
- W2892058377 hasRelatedWork W3026847016 @default.
- W2892058377 hasRelatedWork W3041783013 @default.
- W2892058377 hasRelatedWork W3083215875 @default.
- W2892058377 hasRelatedWork W3096501545 @default.
- W2892058377 hasRelatedWork W3098100992 @default.
- W2892058377 hasRelatedWork W3119830102 @default.
- W2892058377 hasRelatedWork W3173636389 @default.
- W2892058377 hasRelatedWork W3184000328 @default.
- W2892058377 hasRelatedWork W343139615 @default.
- W2892058377 isParatext "false" @default.
- W2892058377 isRetracted "false" @default.
- W2892058377 magId "2892058377" @default.
- W2892058377 workType "dissertation" @default.