Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311404245> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W4311404245 endingPage "11" @default.
- W4311404245 startingPage "1" @default.
- W4311404245 abstract "Unsupervised domain adaptation (UDA) aims to transfer knowledge from a well-labeled source domain to a different but related unlabeled target domain with identical label space. Currently, the main workhorse for solving UDA is domain alignment, which has proven successful. However, it is often difficult to find an appropriate source domain with identical label space. A more practical scenario is so-called partial domain adaptation (PDA) in which the source label set or space subsumes the target one. Unfortunately, in PDA, due to the existence of the irrelevant categories in the source domain, it is quite hard to obtain a perfect alignment, thus resulting in mode collapse and negative transfer. Although several efforts have been made by down-weighting the irrelevant source categories, the strategies used tend to be burdensome and risky since exactly which irrelevant categories are unknown. These challenges motivate us to find a relatively simpler alternative to solve PDA. To achieve this, we first provide a thorough theoretical analysis, which illustrates that the target risk is bounded by both model smoothness and between-domain discrepancy. Considering the difficulty of perfect alignment in solving PDA, we turn to focus on the model smoothness while discard the riskier domain alignment to enhance the adaptability of the model. Specifically, we instantiate the model smoothness as a quite simple intra-domain structure preserving (IDSP). To our best knowledge, this is the first naive attempt to address the PDA without domain alignment. Finally, our empirical results on multiple benchmark datasets demonstrate that IDSP is not only superior to the PDA SOTAs by a significant margin on some benchmarks (e.g., +10% on Cl->Rw and +8% on Ar->Rw ), but also complementary to domain alignment in the standard UDA" @default.
- W4311404245 created "2022-12-26" @default.
- W4311404245 creator A5029012645 @default.
- W4311404245 creator A5047032699 @default.
- W4311404245 date "2022-01-01" @default.
- W4311404245 modified "2023-09-26" @default.
- W4311404245 title "Partial Domain Adaptation without Domain Alignment" @default.
- W4311404245 doi "https://doi.org/10.1109/tpami.2022.3228937" @default.
- W4311404245 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37015373" @default.
- W4311404245 hasPublicationYear "2022" @default.
- W4311404245 type Work @default.
- W4311404245 citedByCount "1" @default.
- W4311404245 countsByYear W43114042452022 @default.
- W4311404245 crossrefType "journal-article" @default.
- W4311404245 hasAuthorship W4311404245A5029012645 @default.
- W4311404245 hasAuthorship W4311404245A5047032699 @default.
- W4311404245 hasBestOaLocation W43114042452 @default.
- W4311404245 hasConcept C102634674 @default.
- W4311404245 hasConcept C111919701 @default.
- W4311404245 hasConcept C11413529 @default.
- W4311404245 hasConcept C114614502 @default.
- W4311404245 hasConcept C134306372 @default.
- W4311404245 hasConcept C153180895 @default.
- W4311404245 hasConcept C154945302 @default.
- W4311404245 hasConcept C184720557 @default.
- W4311404245 hasConcept C2776434776 @default.
- W4311404245 hasConcept C2778572836 @default.
- W4311404245 hasConcept C33923547 @default.
- W4311404245 hasConcept C34388435 @default.
- W4311404245 hasConcept C36503486 @default.
- W4311404245 hasConcept C41008148 @default.
- W4311404245 hasConcept C95623464 @default.
- W4311404245 hasConceptScore W4311404245C102634674 @default.
- W4311404245 hasConceptScore W4311404245C111919701 @default.
- W4311404245 hasConceptScore W4311404245C11413529 @default.
- W4311404245 hasConceptScore W4311404245C114614502 @default.
- W4311404245 hasConceptScore W4311404245C134306372 @default.
- W4311404245 hasConceptScore W4311404245C153180895 @default.
- W4311404245 hasConceptScore W4311404245C154945302 @default.
- W4311404245 hasConceptScore W4311404245C184720557 @default.
- W4311404245 hasConceptScore W4311404245C2776434776 @default.
- W4311404245 hasConceptScore W4311404245C2778572836 @default.
- W4311404245 hasConceptScore W4311404245C33923547 @default.
- W4311404245 hasConceptScore W4311404245C34388435 @default.
- W4311404245 hasConceptScore W4311404245C36503486 @default.
- W4311404245 hasConceptScore W4311404245C41008148 @default.
- W4311404245 hasConceptScore W4311404245C95623464 @default.
- W4311404245 hasLocation W43114042451 @default.
- W4311404245 hasLocation W43114042452 @default.
- W4311404245 hasLocation W43114042453 @default.
- W4311404245 hasOpenAccess W4311404245 @default.
- W4311404245 hasPrimaryLocation W43114042451 @default.
- W4311404245 hasRelatedWork W2322959387 @default.
- W4311404245 hasRelatedWork W2488651947 @default.
- W4311404245 hasRelatedWork W2804535946 @default.
- W4311404245 hasRelatedWork W2949625558 @default.
- W4311404245 hasRelatedWork W3034929501 @default.
- W4311404245 hasRelatedWork W4300223372 @default.
- W4311404245 hasRelatedWork W4300488963 @default.
- W4311404245 hasRelatedWork W4300637975 @default.
- W4311404245 hasRelatedWork W4301497500 @default.
- W4311404245 hasRelatedWork W2986726523 @default.
- W4311404245 isParatext "false" @default.
- W4311404245 isRetracted "false" @default.
- W4311404245 workType "article" @default.