Matches in SemOpenAlex for { <https://semopenalex.org/work/W2997128044> ?p ?o ?g. }
- W2997128044 abstract "Nowadays, deep learning methods, especially the convolutional neural networks (CNNs), have shown impressive performance on extracting abstract and high-level features from the hyperspectral image. However, general training process of CNNs mainly considers the pixel-wise information or the samples' correlation to formulate the penalization while ignores the statistical properties especially the spectral variability of each class in the hyperspectral image. These samples-based penalizations would lead to the uncertainty of the training process due to the imbalanced and limited number of training samples. To overcome this problem, this work characterizes each class from the hyperspectral image as a statistical distribution and further develops a novel statistical loss with the distributions, not directly with samples for deep learning. Based on the Fisher discrimination criterion, the loss penalizes the sample variance of each class distribution to decrease the intra-class variance of the training samples. Moreover, an additional diversity-promoting condition is added to enlarge the inter-class variance between different class distributions and this could better discriminate samples from different classes in hyperspectral image. Finally, the statistical estimation form of the statistical loss is developed with the training samples through multi-variant statistical analysis. Experiments over the real-world hyperspectral images show the effectiveness of the developed statistical loss for deep learning." @default.
- W2997128044 created "2020-01-10" @default.
- W2997128044 creator A5028602988 @default.
- W2997128044 creator A5074772585 @default.
- W2997128044 creator A5079556851 @default.
- W2997128044 date "2019-12-27" @default.
- W2997128044 modified "2023-10-15" @default.
- W2997128044 title "Statistical Loss and Analysis for Deep Learning in Hyperspectral Image Classification" @default.
- W2997128044 cites W1569512666 @default.
- W2997128044 cites W1899348529 @default.
- W2997128044 cites W2017014096 @default.
- W2997128044 cites W2021354639 @default.
- W2997128044 cites W2044184146 @default.
- W2997128044 cites W2090480664 @default.
- W2997128044 cites W2112440119 @default.
- W2997128044 cites W2118996198 @default.
- W2997128044 cites W2138621090 @default.
- W2997128044 cites W2145224695 @default.
- W2997128044 cites W2155893237 @default.
- W2997128044 cites W2157785665 @default.
- W2997128044 cites W2194775991 @default.
- W2997128044 cites W2316226477 @default.
- W2997128044 cites W2417947228 @default.
- W2997128044 cites W2418115551 @default.
- W2997128044 cites W2500751094 @default.
- W2997128044 cites W2516506474 @default.
- W2997128044 cites W2520774990 @default.
- W2997128044 cites W2533102868 @default.
- W2997128044 cites W2548791488 @default.
- W2997128044 cites W2590856740 @default.
- W2997128044 cites W2603422184 @default.
- W2997128044 cites W2621189286 @default.
- W2997128044 cites W2732412926 @default.
- W2997128044 cites W2761385227 @default.
- W2997128044 cites W2764276316 @default.
- W2997128044 cites W2767581044 @default.
- W2997128044 cites W2772452219 @default.
- W2997128044 cites W2792332881 @default.
- W2997128044 cites W2809635958 @default.
- W2997128044 cites W2887785636 @default.
- W2997128044 cites W2907100627 @default.
- W2997128044 cites W2912961521 @default.
- W2997128044 cites W2963026686 @default.
- W2997128044 cites W2963988212 @default.
- W2997128044 cites W3103753223 @default.
- W2997128044 doi "https://doi.org/10.48550/arxiv.1912.12385" @default.
- W2997128044 hasPublicationYear "2019" @default.
- W2997128044 type Work @default.
- W2997128044 sameAs 2997128044 @default.
- W2997128044 citedByCount "0" @default.
- W2997128044 crossrefType "posted-content" @default.
- W2997128044 hasAuthorship W2997128044A5028602988 @default.
- W2997128044 hasAuthorship W2997128044A5074772585 @default.
- W2997128044 hasAuthorship W2997128044A5079556851 @default.
- W2997128044 hasBestOaLocation W29971280441 @default.
- W2997128044 hasConcept C105795698 @default.
- W2997128044 hasConcept C108583219 @default.
- W2997128044 hasConcept C114289077 @default.
- W2997128044 hasConcept C115961682 @default.
- W2997128044 hasConcept C119857082 @default.
- W2997128044 hasConcept C121955636 @default.
- W2997128044 hasConcept C144133560 @default.
- W2997128044 hasConcept C153180895 @default.
- W2997128044 hasConcept C154945302 @default.
- W2997128044 hasConcept C159078339 @default.
- W2997128044 hasConcept C160633673 @default.
- W2997128044 hasConcept C196083921 @default.
- W2997128044 hasConcept C2777212361 @default.
- W2997128044 hasConcept C33923547 @default.
- W2997128044 hasConcept C41008148 @default.
- W2997128044 hasConcept C50644808 @default.
- W2997128044 hasConcept C81363708 @default.
- W2997128044 hasConceptScore W2997128044C105795698 @default.
- W2997128044 hasConceptScore W2997128044C108583219 @default.
- W2997128044 hasConceptScore W2997128044C114289077 @default.
- W2997128044 hasConceptScore W2997128044C115961682 @default.
- W2997128044 hasConceptScore W2997128044C119857082 @default.
- W2997128044 hasConceptScore W2997128044C121955636 @default.
- W2997128044 hasConceptScore W2997128044C144133560 @default.
- W2997128044 hasConceptScore W2997128044C153180895 @default.
- W2997128044 hasConceptScore W2997128044C154945302 @default.
- W2997128044 hasConceptScore W2997128044C159078339 @default.
- W2997128044 hasConceptScore W2997128044C160633673 @default.
- W2997128044 hasConceptScore W2997128044C196083921 @default.
- W2997128044 hasConceptScore W2997128044C2777212361 @default.
- W2997128044 hasConceptScore W2997128044C33923547 @default.
- W2997128044 hasConceptScore W2997128044C41008148 @default.
- W2997128044 hasConceptScore W2997128044C50644808 @default.
- W2997128044 hasConceptScore W2997128044C81363708 @default.
- W2997128044 hasLocation W29971280441 @default.
- W2997128044 hasLocation W29971280442 @default.
- W2997128044 hasOpenAccess W2997128044 @default.
- W2997128044 hasPrimaryLocation W29971280441 @default.
- W2997128044 hasRelatedWork W2019190440 @default.
- W2997128044 hasRelatedWork W2027399350 @default.
- W2997128044 hasRelatedWork W2044184146 @default.
- W2997128044 hasRelatedWork W2060875994 @default.
- W2997128044 hasRelatedWork W2067727414 @default.
- W2997128044 hasRelatedWork W2070598848 @default.
- W2997128044 hasRelatedWork W2072166414 @default.