Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897712637> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W2897712637 abstract "Recently Convolutional Neural Networks(CNNs) models have achieved remarkable results for fine-grained image classification. However, CNNs require a large amount of training data during supervised learning and labeling so much data is expensive in many cases. To address this issue, this paper innovatively presents a semi-supervised pipeline to improve fine-grained classification tasks without any extra data. We carefully combine CNNs with Generative Adversarial Nets(GANs) for classification, which shows that the result is affirmative. In addition, we propose a multi dimension label regularization(MDLR) method to train labeled images and unlabeled images simultaneously. First we use a pre-trained Yolo v2 object detection model to detect coarse-grained object on the original dataset. Second we feed cropped images to the generator of GAN to produce more generated data and assign a uniform label distribution to the generated images. Third we mix these origin real images and generated images. Then these mixed images are fed to a baseline CNN classifier and a feature-fused CNN classifier. We obtain competitive or state-of-the-art results: using feature-fused CNN model on Stanford Dogs dataset we set a new state-of-the-art result of 90.7%; on Oxford 102 Flowers dataset, we show consistent improvements over baseline." @default.
- W2897712637 created "2018-10-26" @default.
- W2897712637 creator A5013878126 @default.
- W2897712637 creator A5032279676 @default.
- W2897712637 creator A5041031140 @default.
- W2897712637 date "2018-09-01" @default.
- W2897712637 modified "2023-09-24" @default.
- W2897712637 title "Improving Fine-Grained Object Classification Using Adversarial Generated Unlabelled Samples" @default.
- W2897712637 cites W1536680647 @default.
- W2897712637 cites W1724369340 @default.
- W2897712637 cites W1928906481 @default.
- W2897712637 cites W1995543189 @default.
- W2897712637 cites W2014102544 @default.
- W2897712637 cites W2014930414 @default.
- W2897712637 cites W2062118960 @default.
- W2897712637 cites W2099471712 @default.
- W2897712637 cites W2110015572 @default.
- W2897712637 cites W2110765924 @default.
- W2897712637 cites W2152411181 @default.
- W2897712637 cites W2173520492 @default.
- W2897712637 cites W2243391935 @default.
- W2897712637 cites W2287418003 @default.
- W2897712637 cites W2306952455 @default.
- W2897712637 cites W2533598788 @default.
- W2897712637 cites W2570343428 @default.
- W2897712637 cites W2585635281 @default.
- W2897712637 cites W2591924527 @default.
- W2897712637 cites W2734819869 @default.
- W2897712637 cites W3106250896 @default.
- W2897712637 cites W3124951096 @default.
- W2897712637 cites W2159423002 @default.
- W2897712637 doi "https://doi.org/10.1109/bigmm.2018.8499075" @default.
- W2897712637 hasPublicationYear "2018" @default.
- W2897712637 type Work @default.
- W2897712637 sameAs 2897712637 @default.
- W2897712637 citedByCount "0" @default.
- W2897712637 crossrefType "proceedings-article" @default.
- W2897712637 hasAuthorship W2897712637A5013878126 @default.
- W2897712637 hasAuthorship W2897712637A5032279676 @default.
- W2897712637 hasAuthorship W2897712637A5041031140 @default.
- W2897712637 hasConcept C108583219 @default.
- W2897712637 hasConcept C115961682 @default.
- W2897712637 hasConcept C153180895 @default.
- W2897712637 hasConcept C154945302 @default.
- W2897712637 hasConcept C2776135515 @default.
- W2897712637 hasConcept C2776151529 @default.
- W2897712637 hasConcept C41008148 @default.
- W2897712637 hasConcept C75294576 @default.
- W2897712637 hasConcept C81363708 @default.
- W2897712637 hasConcept C95623464 @default.
- W2897712637 hasConceptScore W2897712637C108583219 @default.
- W2897712637 hasConceptScore W2897712637C115961682 @default.
- W2897712637 hasConceptScore W2897712637C153180895 @default.
- W2897712637 hasConceptScore W2897712637C154945302 @default.
- W2897712637 hasConceptScore W2897712637C2776135515 @default.
- W2897712637 hasConceptScore W2897712637C2776151529 @default.
- W2897712637 hasConceptScore W2897712637C41008148 @default.
- W2897712637 hasConceptScore W2897712637C75294576 @default.
- W2897712637 hasConceptScore W2897712637C81363708 @default.
- W2897712637 hasConceptScore W2897712637C95623464 @default.
- W2897712637 hasLocation W28977126371 @default.
- W2897712637 hasOpenAccess W2897712637 @default.
- W2897712637 hasPrimaryLocation W28977126371 @default.
- W2897712637 hasRelatedWork W2572787276 @default.
- W2897712637 hasRelatedWork W2773120646 @default.
- W2897712637 hasRelatedWork W2774265021 @default.
- W2897712637 hasRelatedWork W2897712637 @default.
- W2897712637 hasRelatedWork W2920938200 @default.
- W2897712637 hasRelatedWork W2979570315 @default.
- W2897712637 hasRelatedWork W3208464355 @default.
- W2897712637 hasRelatedWork W4200023151 @default.
- W2897712637 hasRelatedWork W4200390792 @default.
- W2897712637 hasRelatedWork W564581980 @default.
- W2897712637 isParatext "false" @default.
- W2897712637 isRetracted "false" @default.
- W2897712637 magId "2897712637" @default.
- W2897712637 workType "article" @default.