Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387592114> ?p ?o ?g. }
- W4387592114 endingPage "108966" @default.
- W4387592114 startingPage "108966" @default.
- W4387592114 abstract "Recently, active domain adaptation (ADA) has become an emerging paradigm, as it significantly boosts the performance of the model with additional supervision. Despite the impressive performance of ADA methods, they are assumed that the source domain can be available during model adaptation. In many scenarios, the source domain can be unavailable due to issues such as data privacy. Consequently, ADA methods fail to achieve the expected results or even become invalid. To solve the problem, we design an effective framework named Feature Mixing and Self-Training (FMAS) for source-free active domain adaptation. Specifically, we adopt Clustering-based Alignment to make the classification boundary clearer to be beneficial to the subsequent sample selection, which can also achieve initial domain adaptation. When reaching active sample selection, we utilize feature mixing to explore the uncertainty and diversity of the sample. We construct the sample candidate pool by Mixup-based Uncertainty Selection and then consider the diversity to overcome the redundancy among the pool by Entropy-based Diversity Selection. Finally, for the remaining unlabeled samples, we adopt the self-training framework, which screens out high-confidence samples to explore the information of unlabeled samples. Thorough experiments show that FMAS improves average accuracy by at least 1% compared to other methods on benchmark datasets OfficeHome and MiniDomainNet. And FMAS can also achieve better results with ADA methods without the source domain data, which demonstrates FMAS can outperform previous ADA methods." @default.
- W4387592114 created "2023-10-13" @default.
- W4387592114 creator A5034227844 @default.
- W4387592114 creator A5037443625 @default.
- W4387592114 date "2023-11-01" @default.
- W4387592114 modified "2023-10-14" @default.
- W4387592114 title "Feature mixing and self-training for source-free active domain adaptation" @default.
- W4387592114 cites W2096943734 @default.
- W4387592114 cites W2120887445 @default.
- W4387592114 cites W2164943005 @default.
- W4387592114 cites W2194775991 @default.
- W4387592114 cites W2412782625 @default.
- W4387592114 cites W2618530766 @default.
- W4387592114 cites W2627183927 @default.
- W4387592114 cites W2956371155 @default.
- W4387592114 cites W2963107255 @default.
- W4387592114 cites W2963446712 @default.
- W4387592114 cites W2964117661 @default.
- W4387592114 cites W2964309882 @default.
- W4387592114 cites W2979509742 @default.
- W4387592114 cites W2980096013 @default.
- W4387592114 cites W2981720610 @default.
- W4387592114 cites W2984353870 @default.
- W4387592114 cites W2986514296 @default.
- W4387592114 cites W2998043922 @default.
- W4387592114 cites W3010360081 @default.
- W4387592114 cites W3014847672 @default.
- W4387592114 cites W3034218934 @default.
- W4387592114 cites W3035499919 @default.
- W4387592114 cites W3173679439 @default.
- W4387592114 cites W3174142717 @default.
- W4387592114 cites W3175042385 @default.
- W4387592114 cites W3182214289 @default.
- W4387592114 cites W3198430326 @default.
- W4387592114 cites W3201074629 @default.
- W4387592114 cites W3202165894 @default.
- W4387592114 cites W3216480618 @default.
- W4387592114 cites W4225162032 @default.
- W4387592114 cites W4304086873 @default.
- W4387592114 cites W4312399508 @default.
- W4387592114 cites W4312677475 @default.
- W4387592114 doi "https://doi.org/10.1016/j.compeleceng.2023.108966" @default.
- W4387592114 hasPublicationYear "2023" @default.
- W4387592114 type Work @default.
- W4387592114 citedByCount "0" @default.
- W4387592114 crossrefType "journal-article" @default.
- W4387592114 hasAuthorship W4387592114A5034227844 @default.
- W4387592114 hasAuthorship W4387592114A5037443625 @default.
- W4387592114 hasConcept C106301342 @default.
- W4387592114 hasConcept C111919701 @default.
- W4387592114 hasConcept C119857082 @default.
- W4387592114 hasConcept C120665830 @default.
- W4387592114 hasConcept C121332964 @default.
- W4387592114 hasConcept C124101348 @default.
- W4387592114 hasConcept C13280743 @default.
- W4387592114 hasConcept C134306372 @default.
- W4387592114 hasConcept C139807058 @default.
- W4387592114 hasConcept C148483581 @default.
- W4387592114 hasConcept C152124472 @default.
- W4387592114 hasConcept C154945302 @default.
- W4387592114 hasConcept C185592680 @default.
- W4387592114 hasConcept C185798385 @default.
- W4387592114 hasConcept C198531522 @default.
- W4387592114 hasConcept C205649164 @default.
- W4387592114 hasConcept C2776434776 @default.
- W4387592114 hasConcept C33923547 @default.
- W4387592114 hasConcept C36503486 @default.
- W4387592114 hasConcept C41008148 @default.
- W4387592114 hasConcept C43617362 @default.
- W4387592114 hasConcept C62520636 @default.
- W4387592114 hasConcept C73555534 @default.
- W4387592114 hasConcept C95623464 @default.
- W4387592114 hasConceptScore W4387592114C106301342 @default.
- W4387592114 hasConceptScore W4387592114C111919701 @default.
- W4387592114 hasConceptScore W4387592114C119857082 @default.
- W4387592114 hasConceptScore W4387592114C120665830 @default.
- W4387592114 hasConceptScore W4387592114C121332964 @default.
- W4387592114 hasConceptScore W4387592114C124101348 @default.
- W4387592114 hasConceptScore W4387592114C13280743 @default.
- W4387592114 hasConceptScore W4387592114C134306372 @default.
- W4387592114 hasConceptScore W4387592114C139807058 @default.
- W4387592114 hasConceptScore W4387592114C148483581 @default.
- W4387592114 hasConceptScore W4387592114C152124472 @default.
- W4387592114 hasConceptScore W4387592114C154945302 @default.
- W4387592114 hasConceptScore W4387592114C185592680 @default.
- W4387592114 hasConceptScore W4387592114C185798385 @default.
- W4387592114 hasConceptScore W4387592114C198531522 @default.
- W4387592114 hasConceptScore W4387592114C205649164 @default.
- W4387592114 hasConceptScore W4387592114C2776434776 @default.
- W4387592114 hasConceptScore W4387592114C33923547 @default.
- W4387592114 hasConceptScore W4387592114C36503486 @default.
- W4387592114 hasConceptScore W4387592114C41008148 @default.
- W4387592114 hasConceptScore W4387592114C43617362 @default.
- W4387592114 hasConceptScore W4387592114C62520636 @default.
- W4387592114 hasConceptScore W4387592114C73555534 @default.
- W4387592114 hasConceptScore W4387592114C95623464 @default.
- W4387592114 hasLocation W43875921141 @default.
- W4387592114 hasOpenAccess W4387592114 @default.