DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lee, Young-Jun | ko |
dc.contributor.author | Choi, Ho-Jin | ko |
dc.date.accessioned | 2023-06-23T02:01:02Z | - |
dc.date.available | 2023-06-23T02:01:02Z | - |
dc.date.created | 2023-06-22 | - |
dc.date.issued | 2023-02-13 | - |
dc.identifier.citation | 2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023, pp.410 - 412 | - |
dc.identifier.uri | http://hdl.handle.net/10203/308692 | - |
dc.description.abstract | Recently, many studies have constructed multimodal dialogue datasets containing image-sharing behavior, which is vital to increase the social relationship with interlocutors in open-domain conversation. In this paper, we report the empirical results that CLIP can understand the alignment between the dialogue history and image by conducting various experiments for (1) zero-shot transferability, (2) the effect of dialogue history, and (3) robustness. Our experiments demonstrate that it is necessary for improving the zero-shot performance of CLIP on the multi-modal dialogue dataset. Additionally, the CLIP model is benefitted from more informative texts (i.e., dialogue history), not the last utterance only. | - |
dc.language | English | - |
dc.publisher | Institute of Electrical and Electronics Engineers Inc. | - |
dc.title | Can CLIP Share Image in Dialogue? | - |
dc.type | Conference | - |
dc.identifier.scopusid | 2-s2.0-85151524327 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 410 | - |
dc.citation.endingpage | 412 | - |
dc.citation.publicationname | 2023 IEEE International Conference on Big Data and Smart Computing, BigComp 2023 | - |
dc.identifier.conferencecountry | KO | - |
dc.identifier.conferencelocation | Jeju Island | - |
dc.identifier.doi | 10.1109/BigComp57234.2023.00101 | - |
dc.contributor.localauthor | Choi, Ho-Jin | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.