Matches in SemOpenAlex for { <https://semopenalex.org/work/W4205430752> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W4205430752 endingPage "1220" @default.
- W4205430752 startingPage "1210" @default.
- W4205430752 abstract "Adapting deep neural networks to the changing environments is critical in practical utility, especially for online web applications, where the data distribution changes gradually due to the evolving environments. For instance, the web photos of cellphones change gradually over years due to appearance changes. This paper deals with such a problem via active gradual domain adaptation, where the learner continually and actively selects the most informative labels from the target to enhance labeling efficiency and utilizes both labeled and unlabeled samples to improve the model adaptation under gradual domain drift. We propose the active gradual self-training (AGST) algorithm with novel designs of active pseudolabeling and gradual semi-supervised domain adaptation. Specifically, AGST pseudolabels the samples with high confidence, and selects the most informative labels from the unconfident samples based on both uncertainty and diversity, and then gradually self-trains itself by confident pseudolabels and queried labels. To study the gradual domain shift problem in the web data and verify the proposed algorithm, we create a new dataset -- Evolving-Image-Search (EVIS), collected from the web search engine and covers a 12-years range. Since the appearance of the products evolves over these years, such dataset naturally contains gradual domain drift. We extensively evaluate AGST on the synthetic dataset, real-world dataset, and EVIS dataset. AGST achieves up to 62% accuracy improvement (absolute value) against unsupervised gradual self-training with only 5% additional labels, and 19% accuracy improvement against directly applying CLUE, demonstrating the effectiveness of the designs of active pseudolabel and gradual semi-supervised domain adaptation." @default.
- W4205430752 created "2022-01-26" @default.
- W4205430752 creator A5013030532 @default.
- W4205430752 creator A5014011362 @default.
- W4205430752 creator A5036016529 @default.
- W4205430752 creator A5079077671 @default.
- W4205430752 creator A5086273618 @default.
- W4205430752 date "2022-01-01" @default.
- W4205430752 modified "2023-10-13" @default.
- W4205430752 title "Active Gradual Domain Adaptation: Dataset and Approach" @default.
- W4205430752 cites W1536680647 @default.
- W4205430752 cites W2034368206 @default.
- W4205430752 cites W2135148528 @default.
- W4205430752 cites W2163922914 @default.
- W4205430752 cites W2194775991 @default.
- W4205430752 cites W2334687549 @default.
- W4205430752 cites W2557728737 @default.
- W4205430752 cites W2593768305 @default.
- W4205430752 cites W2627183927 @default.
- W4205430752 cites W2756073160 @default.
- W4205430752 cites W2786808285 @default.
- W4205430752 cites W2963826694 @default.
- W4205430752 cites W2964126011 @default.
- W4205430752 cites W2984810221 @default.
- W4205430752 cites W2986381065 @default.
- W4205430752 cites W3010360081 @default.
- W4205430752 cites W3094277917 @default.
- W4205430752 cites W3102842772 @default.
- W4205430752 cites W3202165894 @default.
- W4205430752 cites W4230030242 @default.
- W4205430752 doi "https://doi.org/10.1109/tmm.2022.3142524" @default.
- W4205430752 hasPublicationYear "2022" @default.
- W4205430752 type Work @default.
- W4205430752 citedByCount "5" @default.
- W4205430752 countsByYear W42054307522023 @default.
- W4205430752 crossrefType "journal-article" @default.
- W4205430752 hasAuthorship W4205430752A5013030532 @default.
- W4205430752 hasAuthorship W4205430752A5014011362 @default.
- W4205430752 hasAuthorship W4205430752A5036016529 @default.
- W4205430752 hasAuthorship W4205430752A5079077671 @default.
- W4205430752 hasAuthorship W4205430752A5086273618 @default.
- W4205430752 hasConcept C119857082 @default.
- W4205430752 hasConcept C120665830 @default.
- W4205430752 hasConcept C121332964 @default.
- W4205430752 hasConcept C124101348 @default.
- W4205430752 hasConcept C134306372 @default.
- W4205430752 hasConcept C139807058 @default.
- W4205430752 hasConcept C153180895 @default.
- W4205430752 hasConcept C154945302 @default.
- W4205430752 hasConcept C159985019 @default.
- W4205430752 hasConcept C192562407 @default.
- W4205430752 hasConcept C204323151 @default.
- W4205430752 hasConcept C2776434776 @default.
- W4205430752 hasConcept C33923547 @default.
- W4205430752 hasConcept C36503486 @default.
- W4205430752 hasConcept C41008148 @default.
- W4205430752 hasConcept C95623464 @default.
- W4205430752 hasConceptScore W4205430752C119857082 @default.
- W4205430752 hasConceptScore W4205430752C120665830 @default.
- W4205430752 hasConceptScore W4205430752C121332964 @default.
- W4205430752 hasConceptScore W4205430752C124101348 @default.
- W4205430752 hasConceptScore W4205430752C134306372 @default.
- W4205430752 hasConceptScore W4205430752C139807058 @default.
- W4205430752 hasConceptScore W4205430752C153180895 @default.
- W4205430752 hasConceptScore W4205430752C154945302 @default.
- W4205430752 hasConceptScore W4205430752C159985019 @default.
- W4205430752 hasConceptScore W4205430752C192562407 @default.
- W4205430752 hasConceptScore W4205430752C204323151 @default.
- W4205430752 hasConceptScore W4205430752C2776434776 @default.
- W4205430752 hasConceptScore W4205430752C33923547 @default.
- W4205430752 hasConceptScore W4205430752C36503486 @default.
- W4205430752 hasConceptScore W4205430752C41008148 @default.
- W4205430752 hasConceptScore W4205430752C95623464 @default.
- W4205430752 hasFunder F4320321001 @default.
- W4205430752 hasFunder F4320326705 @default.
- W4205430752 hasLocation W42054307521 @default.
- W4205430752 hasOpenAccess W4205430752 @default.
- W4205430752 hasPrimaryLocation W42054307521 @default.
- W4205430752 hasRelatedWork W1864954421 @default.
- W4205430752 hasRelatedWork W2152148513 @default.
- W4205430752 hasRelatedWork W2280198878 @default.
- W4205430752 hasRelatedWork W2783380393 @default.
- W4205430752 hasRelatedWork W2951876757 @default.
- W4205430752 hasRelatedWork W3012432583 @default.
- W4205430752 hasRelatedWork W3096565154 @default.
- W4205430752 hasRelatedWork W4281397339 @default.
- W4205430752 hasRelatedWork W4283450023 @default.
- W4205430752 hasRelatedWork W4312910505 @default.
- W4205430752 hasVolume "24" @default.
- W4205430752 isParatext "false" @default.
- W4205430752 isRetracted "false" @default.
- W4205430752 workType "article" @default.