Matches in SemOpenAlex for { <https://semopenalex.org/work/W4301356211> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W4301356211 abstract "We introduce the first goal-driven training for visual question answering and dialog agents. Specifically, we pose a cooperative 'image guessing' game between two agents -- Qbot and Abot -- who communicate in natural language dialog so that Qbot can select an unseen image from a lineup of images. We use deep reinforcement learning (RL) to learn the policies of these agents end-to-end -- from pixels to multi-agent multi-round dialog to game reward. We demonstrate two experimental results. First, as a 'sanity check' demonstration of pure RL (from scratch), we show results on a synthetic world, where the agents communicate in ungrounded vocabulary, i.e., symbols with no pre-specified meanings (X, Y, Z). We find that two bots invent their own communication protocol and start using certain symbols to ask/answer about certain visual attributes (shape/color/style). Thus, we demonstrate the emergence of grounded language and communication among 'visual' dialog agents with no human supervision. Second, we conduct large-scale real-image experiments on the VisDial dataset, where we pretrain with supervised dialog data and show that the RL 'fine-tuned' agents significantly outperform SL agents. Interestingly, the RL Qbot learns to ask questions that Abot is good at, ultimately resulting in more informative dialog and a better team." @default.
- W4301356211 created "2022-10-05" @default.
- W4301356211 creator A5014035752 @default.
- W4301356211 creator A5045861415 @default.
- W4301356211 creator A5051259505 @default.
- W4301356211 creator A5070914582 @default.
- W4301356211 creator A5088655875 @default.
- W4301356211 date "2017-03-19" @default.
- W4301356211 modified "2023-09-28" @default.
- W4301356211 title "Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning" @default.
- W4301356211 doi "https://doi.org/10.48550/arxiv.1703.06585" @default.
- W4301356211 hasPublicationYear "2017" @default.
- W4301356211 type Work @default.
- W4301356211 citedByCount "0" @default.
- W4301356211 crossrefType "posted-content" @default.
- W4301356211 hasAuthorship W4301356211A5014035752 @default.
- W4301356211 hasAuthorship W4301356211A5045861415 @default.
- W4301356211 hasAuthorship W4301356211A5051259505 @default.
- W4301356211 hasAuthorship W4301356211A5070914582 @default.
- W4301356211 hasAuthorship W4301356211A5088655875 @default.
- W4301356211 hasBestOaLocation W43013562111 @default.
- W4301356211 hasConcept C107457646 @default.
- W4301356211 hasConcept C136264566 @default.
- W4301356211 hasConcept C136764020 @default.
- W4301356211 hasConcept C138885662 @default.
- W4301356211 hasConcept C154945302 @default.
- W4301356211 hasConcept C162324750 @default.
- W4301356211 hasConcept C173853756 @default.
- W4301356211 hasConcept C190954187 @default.
- W4301356211 hasConcept C195324797 @default.
- W4301356211 hasConcept C204321447 @default.
- W4301356211 hasConcept C2777601683 @default.
- W4301356211 hasConcept C41008148 @default.
- W4301356211 hasConcept C41895202 @default.
- W4301356211 hasConcept C90329073 @default.
- W4301356211 hasConcept C97541855 @default.
- W4301356211 hasConceptScore W4301356211C107457646 @default.
- W4301356211 hasConceptScore W4301356211C136264566 @default.
- W4301356211 hasConceptScore W4301356211C136764020 @default.
- W4301356211 hasConceptScore W4301356211C138885662 @default.
- W4301356211 hasConceptScore W4301356211C154945302 @default.
- W4301356211 hasConceptScore W4301356211C162324750 @default.
- W4301356211 hasConceptScore W4301356211C173853756 @default.
- W4301356211 hasConceptScore W4301356211C190954187 @default.
- W4301356211 hasConceptScore W4301356211C195324797 @default.
- W4301356211 hasConceptScore W4301356211C204321447 @default.
- W4301356211 hasConceptScore W4301356211C2777601683 @default.
- W4301356211 hasConceptScore W4301356211C41008148 @default.
- W4301356211 hasConceptScore W4301356211C41895202 @default.
- W4301356211 hasConceptScore W4301356211C90329073 @default.
- W4301356211 hasConceptScore W4301356211C97541855 @default.
- W4301356211 hasLocation W43013562111 @default.
- W4301356211 hasOpenAccess W4301356211 @default.
- W4301356211 hasPrimaryLocation W43013562111 @default.
- W4301356211 hasRelatedWork W1532946172 @default.
- W4301356211 hasRelatedWork W1628671981 @default.
- W4301356211 hasRelatedWork W1767774682 @default.
- W4301356211 hasRelatedWork W1900635591 @default.
- W4301356211 hasRelatedWork W1963944933 @default.
- W4301356211 hasRelatedWork W2013809956 @default.
- W4301356211 hasRelatedWork W2063157598 @default.
- W4301356211 hasRelatedWork W2377424484 @default.
- W4301356211 hasRelatedWork W2553439369 @default.
- W4301356211 hasRelatedWork W1872130062 @default.
- W4301356211 isParatext "false" @default.
- W4301356211 isRetracted "false" @default.
- W4301356211 workType "article" @default.