Matches in SemOpenAlex for { <https://semopenalex.org/work/W3095945950> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W3095945950 endingPage "159" @default.
- W3095945950 startingPage "152" @default.
- W3095945950 abstract "AbstractIn this paper, we present a two-step deep learning technique in support of DDDAS for achievement of robust ATR via transfer learning using simulated SAR imagery. The first Deep Learning (DL) model performs noise suppression of input SAR images via a Multi-resolution Stacked Denoising Autoencoder (MSDAE) architecture. The second DL model includes a Multi-output Convolutional Neural Network (M-CNN) architecture suitable for multi-feature classification of ATR pertaining to the DDDAS paradigm. In this approach, we train each DL model independently, then, streamline this process as a standalone deep learning ATR classifier. Primarily, we employed the IRIS Electromagnetic (IRIS-EM) modeling and simulation system to systematically generate our own multi-look large-scale simulated SAR images of multi-platform (i.e., ground, aerial, and marine) vehicles. To improve situational awareness of a DDDAS with respect to ATR, we devised dynamic transfer learnings which employ a step-wise retraining inspired by the observational statistical sampling technique. In this paper, we demonstrate the efficiency and effectiveness of the proposed approach in performing multi-feature ATR of test target vehicles applicable to DDDAS. Lastly, we discuss our classification results using a streamlined denoising and classification system and justify its implication for the DL-based DDDAS.KeywordsDeep learningSARATRDDDASImage denoisingConvolutional neural networks" @default.
- W3095945950 created "2020-11-09" @default.
- W3095945950 creator A5014238481 @default.
- W3095945950 creator A5017935008 @default.
- W3095945950 creator A5019964738 @default.
- W3095945950 creator A5039105022 @default.
- W3095945950 date "2020-01-01" @default.
- W3095945950 modified "2023-10-18" @default.
- W3095945950 title "Dynamic Transfer Learning from Physics-Based Simulated SAR Imagery for Automatic Target Recognition" @default.
- W3095945950 cites W2025768430 @default.
- W3095945950 cites W2051941592 @default.
- W3095945950 cites W2133665775 @default.
- W3095945950 cites W3017554714 @default.
- W3095945950 cites W3017769335 @default.
- W3095945950 cites W3021909624 @default.
- W3095945950 doi "https://doi.org/10.1007/978-3-030-61725-7_19" @default.
- W3095945950 hasPublicationYear "2020" @default.
- W3095945950 type Work @default.
- W3095945950 sameAs 3095945950 @default.
- W3095945950 citedByCount "0" @default.
- W3095945950 crossrefType "book-chapter" @default.
- W3095945950 hasAuthorship W3095945950A5014238481 @default.
- W3095945950 hasAuthorship W3095945950A5017935008 @default.
- W3095945950 hasAuthorship W3095945950A5019964738 @default.
- W3095945950 hasAuthorship W3095945950A5039105022 @default.
- W3095945950 hasConcept C101738243 @default.
- W3095945950 hasConcept C108583219 @default.
- W3095945950 hasConcept C117623542 @default.
- W3095945950 hasConcept C150899416 @default.
- W3095945950 hasConcept C153180895 @default.
- W3095945950 hasConcept C154945302 @default.
- W3095945950 hasConcept C41008148 @default.
- W3095945950 hasConcept C81363708 @default.
- W3095945950 hasConcept C87360688 @default.
- W3095945950 hasConcept C95623464 @default.
- W3095945950 hasConceptScore W3095945950C101738243 @default.
- W3095945950 hasConceptScore W3095945950C108583219 @default.
- W3095945950 hasConceptScore W3095945950C117623542 @default.
- W3095945950 hasConceptScore W3095945950C150899416 @default.
- W3095945950 hasConceptScore W3095945950C153180895 @default.
- W3095945950 hasConceptScore W3095945950C154945302 @default.
- W3095945950 hasConceptScore W3095945950C41008148 @default.
- W3095945950 hasConceptScore W3095945950C81363708 @default.
- W3095945950 hasConceptScore W3095945950C87360688 @default.
- W3095945950 hasConceptScore W3095945950C95623464 @default.
- W3095945950 hasLocation W30959459501 @default.
- W3095945950 hasOpenAccess W3095945950 @default.
- W3095945950 hasPrimaryLocation W30959459501 @default.
- W3095945950 hasRelatedWork W12793662 @default.
- W3095945950 hasRelatedWork W12829028 @default.
- W3095945950 hasRelatedWork W15119441 @default.
- W3095945950 hasRelatedWork W2582698 @default.
- W3095945950 hasRelatedWork W3968525 @default.
- W3095945950 hasRelatedWork W4947539 @default.
- W3095945950 hasRelatedWork W5454555 @default.
- W3095945950 hasRelatedWork W6680660 @default.
- W3095945950 hasRelatedWork W7303821 @default.
- W3095945950 hasRelatedWork W9190101 @default.
- W3095945950 isParatext "false" @default.
- W3095945950 isRetracted "false" @default.
- W3095945950 magId "3095945950" @default.
- W3095945950 workType "book-chapter" @default.