Matches in SemOpenAlex for { <https://semopenalex.org/work/W4366148251> ?p ?o ?g. }
- W4366148251 endingPage "3652" @default.
- W4366148251 startingPage "3645" @default.
- W4366148251 abstract "Unsupervised domain adaptive object detection is a challenging perception task where object detectors are adapted from a label-rich source domain to an unlabeled target domain, playing a vital role in autonomous driving and robot navigation. Since the camera settings, weather, and light conditions vary depending on the road, it is labor-intensive and impossible to manually annotate all the data to adapt to these scenarios. Recent advances demonstrate the effectiveness of adversarial-based domain alignment, where domain invariance in the feature space is produced by adversarial training between the feature extractor and domain discriminator. However, due to the domain shift, domain discrimination, especially on low-level features, is an easy task. This results in an imbalance of the adversarial training between the domain discriminator and the feature extractor. In this work, we achieve a better domain alignment by introducing an auxiliary regularization task to improve the training balance. Specifically, we propose Adversarial Image Reconstruction (AIR) as the regularizer to facilitate the adversarial training of the feature extractor. We further design a multi-level Feature Alignment(A) module to enhance the adaptation performance. We refer our work as AIRA-DA for ease of presentation. Our evaluations across several datasets of challenging domain shifts demonstrate that the proposed method outperforms all previous methods, of both one- and two-stage, in most settings. Extensive experiments shows that the proposed method is capable of handling weather variation, cross-camera adaptation and synthetic-to-real-world adaptation." @default.
- W4366148251 created "2023-04-19" @default.
- W4366148251 creator A5010643810 @default.
- W4366148251 creator A5011359067 @default.
- W4366148251 creator A5019360899 @default.
- W4366148251 creator A5049765236 @default.
- W4366148251 creator A5050305421 @default.
- W4366148251 creator A5056225611 @default.
- W4366148251 creator A5089042848 @default.
- W4366148251 date "2023-06-01" @default.
- W4366148251 modified "2023-09-24" @default.
- W4366148251 title "AIRA-DA: Adversarial Image Reconstruction Alignments for Unsupervised Domain Adaptive Object Detection" @default.
- W4366148251 cites W2194775991 @default.
- W4366148251 cites W2340897893 @default.
- W4366148251 cites W2478454054 @default.
- W4366148251 cites W2565639579 @default.
- W4366148251 cites W2618530766 @default.
- W4366148251 cites W2748021867 @default.
- W4366148251 cites W2884561390 @default.
- W4366148251 cites W2904631343 @default.
- W4366148251 cites W2955889502 @default.
- W4366148251 cites W2964115968 @default.
- W4366148251 cites W2964166085 @default.
- W4366148251 cites W2968634921 @default.
- W4366148251 cites W2969583814 @default.
- W4366148251 cites W2979548969 @default.
- W4366148251 cites W2982770724 @default.
- W4366148251 cites W2990069979 @default.
- W4366148251 cites W2990740643 @default.
- W4366148251 cites W3012573144 @default.
- W4366148251 cites W3034779842 @default.
- W4366148251 cites W3095753995 @default.
- W4366148251 cites W3096609285 @default.
- W4366148251 cites W3098090606 @default.
- W4366148251 cites W3102057471 @default.
- W4366148251 cites W3106723659 @default.
- W4366148251 cites W3107090469 @default.
- W4366148251 cites W3107331169 @default.
- W4366148251 cites W3107610277 @default.
- W4366148251 cites W3109202794 @default.
- W4366148251 cites W3110011650 @default.
- W4366148251 cites W3176895448 @default.
- W4366148251 cites W639708223 @default.
- W4366148251 doi "https://doi.org/10.1109/lra.2023.3267692" @default.
- W4366148251 hasPublicationYear "2023" @default.
- W4366148251 type Work @default.
- W4366148251 citedByCount "0" @default.
- W4366148251 crossrefType "journal-article" @default.
- W4366148251 hasAuthorship W4366148251A5010643810 @default.
- W4366148251 hasAuthorship W4366148251A5011359067 @default.
- W4366148251 hasAuthorship W4366148251A5019360899 @default.
- W4366148251 hasAuthorship W4366148251A5049765236 @default.
- W4366148251 hasAuthorship W4366148251A5050305421 @default.
- W4366148251 hasAuthorship W4366148251A5056225611 @default.
- W4366148251 hasAuthorship W4366148251A5089042848 @default.
- W4366148251 hasConcept C134306372 @default.
- W4366148251 hasConcept C138885662 @default.
- W4366148251 hasConcept C153180895 @default.
- W4366148251 hasConcept C154945302 @default.
- W4366148251 hasConcept C2776135515 @default.
- W4366148251 hasConcept C2776151529 @default.
- W4366148251 hasConcept C2776401178 @default.
- W4366148251 hasConcept C2779803651 @default.
- W4366148251 hasConcept C31972630 @default.
- W4366148251 hasConcept C33923547 @default.
- W4366148251 hasConcept C36503486 @default.
- W4366148251 hasConcept C41008148 @default.
- W4366148251 hasConcept C41895202 @default.
- W4366148251 hasConcept C52622490 @default.
- W4366148251 hasConcept C76155785 @default.
- W4366148251 hasConcept C94915269 @default.
- W4366148251 hasConceptScore W4366148251C134306372 @default.
- W4366148251 hasConceptScore W4366148251C138885662 @default.
- W4366148251 hasConceptScore W4366148251C153180895 @default.
- W4366148251 hasConceptScore W4366148251C154945302 @default.
- W4366148251 hasConceptScore W4366148251C2776135515 @default.
- W4366148251 hasConceptScore W4366148251C2776151529 @default.
- W4366148251 hasConceptScore W4366148251C2776401178 @default.
- W4366148251 hasConceptScore W4366148251C2779803651 @default.
- W4366148251 hasConceptScore W4366148251C31972630 @default.
- W4366148251 hasConceptScore W4366148251C33923547 @default.
- W4366148251 hasConceptScore W4366148251C36503486 @default.
- W4366148251 hasConceptScore W4366148251C41008148 @default.
- W4366148251 hasConceptScore W4366148251C41895202 @default.
- W4366148251 hasConceptScore W4366148251C52622490 @default.
- W4366148251 hasConceptScore W4366148251C76155785 @default.
- W4366148251 hasConceptScore W4366148251C94915269 @default.
- W4366148251 hasFunder F4320321001 @default.
- W4366148251 hasFunder F4320335787 @default.
- W4366148251 hasIssue "6" @default.
- W4366148251 hasLocation W43661482511 @default.
- W4366148251 hasOpenAccess W4366148251 @default.
- W4366148251 hasPrimaryLocation W43661482511 @default.
- W4366148251 hasRelatedWork W137416770 @default.
- W4366148251 hasRelatedWork W1504288058 @default.
- W4366148251 hasRelatedWork W2016461833 @default.
- W4366148251 hasRelatedWork W2059299633 @default.
- W4366148251 hasRelatedWork W2090093270 @default.