Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313591672> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W4313591672 endingPage "14" @default.
- W4313591672 startingPage "1" @default.
- W4313591672 abstract "Recent works attempt to employ pre-training in Vision-and-Language Navigation (VLN). However, these methods neglect the importance of historical contexts or ignore predicting future actions during pre-training, limiting the learning of visual-textual correspondence and the capability of decision-making. To address these problems, we present a history-enhanced and order-aware pre-training with the complementing fine-tuning paradigm (HOP+) for VLN. Specifically, besides the common Masked Language Modeling (MLM) and Trajectory-Instruction Matching (TIM) tasks, we design three novel VLN-specific proxy tasks: Action Prediction with History (APH) task, Trajectory Order Modeling (TOM) task and Group Order Modeling (GOM) task. APH task takes into account the visual perception trajectory to enhance the learning of historical knowledge as well as action prediction. The two temporal visual-textual alignment tasks, TOM and GOM further improve the agent's ability to order reasoning. Moreover, we design a memory network to address the representation inconsistency of history context between the pre-training and the fine-tuning stages. The memory network effectively selects and summarizes historical information for action prediction during fine-tuning, without costing huge extra computation consumption for downstream VLN tasks. HOP+ achieves new state-of-the-art performance on four downstream VLN tasks (R2R, REVERIE, RxR, and NDH), which demonstrates the effectiveness of our proposed method." @default.
- W4313591672 created "2023-01-06" @default.
- W4313591672 creator A5017394268 @default.
- W4313591672 creator A5031349440 @default.
- W4313591672 creator A5036105537 @default.
- W4313591672 creator A5049868894 @default.
- W4313591672 creator A5060958969 @default.
- W4313591672 creator A5070842891 @default.
- W4313591672 date "2023-01-01" @default.
- W4313591672 modified "2023-09-25" @default.
- W4313591672 title "HOP+: History-Enhanced and Order-Aware Pre-Training for Vision-and-Language Navigation" @default.
- W4313591672 cites W2117539524 @default.
- W4313591672 cites W2194775991 @default.
- W4313591672 cites W2489434015 @default.
- W4313591672 cites W2550821151 @default.
- W4313591672 cites W2886641317 @default.
- W4313591672 cites W2926977875 @default.
- W4313591672 cites W2962744691 @default.
- W4313591672 cites W2962781483 @default.
- W4313591672 cites W2963800628 @default.
- W4313591672 cites W2964935470 @default.
- W4313591672 cites W2967186499 @default.
- W4313591672 cites W2970231061 @default.
- W4313591672 cites W2970340522 @default.
- W4313591672 cites W2974759213 @default.
- W4313591672 cites W2979727876 @default.
- W4313591672 cites W2981799368 @default.
- W4313591672 cites W2987914945 @default.
- W4313591672 cites W3034376488 @default.
- W4313591672 cites W3034500398 @default.
- W4313591672 cites W3034578524 @default.
- W4313591672 cites W3091588028 @default.
- W4313591672 cites W3100923070 @default.
- W4313591672 cites W3106641651 @default.
- W4313591672 cites W3107069568 @default.
- W4313591672 cites W3109085430 @default.
- W4313591672 cites W3109097593 @default.
- W4313591672 cites W3109380382 @default.
- W4313591672 cites W3165915253 @default.
- W4313591672 cites W3170842411 @default.
- W4313591672 cites W3179213173 @default.
- W4313591672 cites W3192009892 @default.
- W4313591672 cites W3195026654 @default.
- W4313591672 cites W3205276578 @default.
- W4313591672 cites W3206064582 @default.
- W4313591672 cites W4214700710 @default.
- W4313591672 doi "https://doi.org/10.1109/tpami.2023.3234243" @default.
- W4313591672 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37018268" @default.
- W4313591672 hasPublicationYear "2023" @default.
- W4313591672 type Work @default.
- W4313591672 citedByCount "2" @default.
- W4313591672 countsByYear W43135916722023 @default.
- W4313591672 crossrefType "journal-article" @default.
- W4313591672 hasAuthorship W4313591672A5017394268 @default.
- W4313591672 hasAuthorship W4313591672A5031349440 @default.
- W4313591672 hasAuthorship W4313591672A5036105537 @default.
- W4313591672 hasAuthorship W4313591672A5049868894 @default.
- W4313591672 hasAuthorship W4313591672A5060958969 @default.
- W4313591672 hasAuthorship W4313591672A5070842891 @default.
- W4313591672 hasConcept C119857082 @default.
- W4313591672 hasConcept C121332964 @default.
- W4313591672 hasConcept C1276947 @default.
- W4313591672 hasConcept C13662910 @default.
- W4313591672 hasConcept C154945302 @default.
- W4313591672 hasConcept C162324750 @default.
- W4313591672 hasConcept C175154964 @default.
- W4313591672 hasConcept C187736073 @default.
- W4313591672 hasConcept C2780451532 @default.
- W4313591672 hasConcept C41008148 @default.
- W4313591672 hasConceptScore W4313591672C119857082 @default.
- W4313591672 hasConceptScore W4313591672C121332964 @default.
- W4313591672 hasConceptScore W4313591672C1276947 @default.
- W4313591672 hasConceptScore W4313591672C13662910 @default.
- W4313591672 hasConceptScore W4313591672C154945302 @default.
- W4313591672 hasConceptScore W4313591672C162324750 @default.
- W4313591672 hasConceptScore W4313591672C175154964 @default.
- W4313591672 hasConceptScore W4313591672C187736073 @default.
- W4313591672 hasConceptScore W4313591672C2780451532 @default.
- W4313591672 hasConceptScore W4313591672C41008148 @default.
- W4313591672 hasLocation W43135916721 @default.
- W4313591672 hasLocation W43135916722 @default.
- W4313591672 hasOpenAccess W4313591672 @default.
- W4313591672 hasPrimaryLocation W43135916721 @default.
- W4313591672 hasRelatedWork W2081647779 @default.
- W4313591672 hasRelatedWork W2961085424 @default.
- W4313591672 hasRelatedWork W3046775127 @default.
- W4313591672 hasRelatedWork W3138568041 @default.
- W4313591672 hasRelatedWork W3170094116 @default.
- W4313591672 hasRelatedWork W4285260836 @default.
- W4313591672 hasRelatedWork W4286629047 @default.
- W4313591672 hasRelatedWork W4306321456 @default.
- W4313591672 hasRelatedWork W4306674287 @default.
- W4313591672 hasRelatedWork W4224009465 @default.
- W4313591672 isParatext "false" @default.
- W4313591672 isRetracted "false" @default.
- W4313591672 workType "article" @default.