Matches in SemOpenAlex for { <https://semopenalex.org/work/W3211887611> ?p ?o ?g. }
- W3211887611 abstract "Communication between human and mobile agents is getting increasingly important as such agents are widely deployed in our daily lives. Vision-and-Dialogue Navigation is one of the tasks that evaluate the agent’s ability to interact with humans for assistance and navigate based on natural language responses. In this paper, we explore the Navigation from Dialogue History (NDH) task, which is based on the Cooperative Vision-and-Dialogue Navigation (CVDN) dataset, and present a state-of-the-art model which is built upon Vision-Language transformers. However, despite achieving competitive performance, we find that the agent in the NDH task is not evaluated appropriately by the primary metric – Goal Progress. By analyzing the performance mismatch between Goal Progress and other metrics (e.g., normalized Dynamic Time Warping) from our state-of-the-art model, we show that NDH’s sub-path based task setup (i.e., navigating partial trajectory based on its correspondent subset of the full dialogue) does not provide the agent with enough supervision signal towards the goal region. Therefore, we propose a new task setup called NDH-Full which takes the full dialogue and the whole navigation path as one instance. We present a strong baseline model and show initial results on this new task. We further describe several approaches that we try, in order to improve the model performance (based on curriculum learning, pre-training, and data-augmentation), suggesting potential useful training methods on this new NDH-Full task." @default.
- W3211887611 created "2021-11-22" @default.
- W3211887611 creator A5001987532 @default.
- W3211887611 creator A5041514597 @default.
- W3211887611 creator A5087164997 @default.
- W3211887611 date "2021-01-01" @default.
- W3211887611 modified "2023-09-23" @default.
- W3211887611 title "NDH-Full: Learning and Evaluating Navigational Agents on Full-Length Dialogue" @default.
- W3211887611 cites W1602500555 @default.
- W3211887611 cites W2118781169 @default.
- W3211887611 cites W2130463046 @default.
- W3211887611 cites W2194775991 @default.
- W3211887611 cites W2236233024 @default.
- W3211887611 cites W2585954273 @default.
- W3211887611 cites W2627585944 @default.
- W3211887611 cites W2774005037 @default.
- W3211887611 cites W2795911278 @default.
- W3211887611 cites W2835434549 @default.
- W3211887611 cites W2890902815 @default.
- W3211887611 cites W2899235916 @default.
- W3211887611 cites W2899771611 @default.
- W3211887611 cites W2908510526 @default.
- W3211887611 cites W2926977875 @default.
- W3211887611 cites W2951973805 @default.
- W3211887611 cites W2958574008 @default.
- W3211887611 cites W2963341956 @default.
- W3211887611 cites W2963726321 @default.
- W3211887611 cites W2964043796 @default.
- W3211887611 cites W2964121744 @default.
- W3211887611 cites W2964217371 @default.
- W3211887611 cites W2967186499 @default.
- W3211887611 cites W2970231061 @default.
- W3211887611 cites W2970340522 @default.
- W3211887611 cites W2970608575 @default.
- W3211887611 cites W2979727876 @default.
- W3211887611 cites W2981851019 @default.
- W3211887611 cites W2997886377 @default.
- W3211887611 cites W3023306062 @default.
- W3211887611 cites W3029418112 @default.
- W3211887611 cites W3030193665 @default.
- W3211887611 cites W3034253961 @default.
- W3211887611 cites W3034376488 @default.
- W3211887611 cites W3034578524 @default.
- W3211887611 cites W3034758614 @default.
- W3211887611 cites W3034984114 @default.
- W3211887611 cites W3090072755 @default.
- W3211887611 cites W3090449556 @default.
- W3211887611 cites W3100923070 @default.
- W3211887611 cites W3101009265 @default.
- W3211887611 cites W3106571068 @default.
- W3211887611 cites W3108144224 @default.
- W3211887611 cites W3135367836 @default.
- W3211887611 cites W3084181639 @default.
- W3211887611 doi "https://doi.org/10.18653/v1/2021.emnlp-main.518" @default.
- W3211887611 hasPublicationYear "2021" @default.
- W3211887611 type Work @default.
- W3211887611 sameAs 3211887611 @default.
- W3211887611 citedByCount "3" @default.
- W3211887611 countsByYear W32118876112022 @default.
- W3211887611 crossrefType "proceedings-article" @default.
- W3211887611 hasAuthorship W3211887611A5001987532 @default.
- W3211887611 hasAuthorship W3211887611A5041514597 @default.
- W3211887611 hasAuthorship W3211887611A5087164997 @default.
- W3211887611 hasBestOaLocation W32118876111 @default.
- W3211887611 hasConcept C107457646 @default.
- W3211887611 hasConcept C111368507 @default.
- W3211887611 hasConcept C11413529 @default.
- W3211887611 hasConcept C119857082 @default.
- W3211887611 hasConcept C121332964 @default.
- W3211887611 hasConcept C12725497 @default.
- W3211887611 hasConcept C127313418 @default.
- W3211887611 hasConcept C127413603 @default.
- W3211887611 hasConcept C1276947 @default.
- W3211887611 hasConcept C13662910 @default.
- W3211887611 hasConcept C154945302 @default.
- W3211887611 hasConcept C176217482 @default.
- W3211887611 hasConcept C201995342 @default.
- W3211887611 hasConcept C21547014 @default.
- W3211887611 hasConcept C2780451532 @default.
- W3211887611 hasConcept C41008148 @default.
- W3211887611 hasConcept C48103436 @default.
- W3211887611 hasConcept C88516994 @default.
- W3211887611 hasConceptScore W3211887611C107457646 @default.
- W3211887611 hasConceptScore W3211887611C111368507 @default.
- W3211887611 hasConceptScore W3211887611C11413529 @default.
- W3211887611 hasConceptScore W3211887611C119857082 @default.
- W3211887611 hasConceptScore W3211887611C121332964 @default.
- W3211887611 hasConceptScore W3211887611C12725497 @default.
- W3211887611 hasConceptScore W3211887611C127313418 @default.
- W3211887611 hasConceptScore W3211887611C127413603 @default.
- W3211887611 hasConceptScore W3211887611C1276947 @default.
- W3211887611 hasConceptScore W3211887611C13662910 @default.
- W3211887611 hasConceptScore W3211887611C154945302 @default.
- W3211887611 hasConceptScore W3211887611C176217482 @default.
- W3211887611 hasConceptScore W3211887611C201995342 @default.
- W3211887611 hasConceptScore W3211887611C21547014 @default.
- W3211887611 hasConceptScore W3211887611C2780451532 @default.
- W3211887611 hasConceptScore W3211887611C41008148 @default.
- W3211887611 hasConceptScore W3211887611C48103436 @default.
- W3211887611 hasConceptScore W3211887611C88516994 @default.