Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Li, Qinbo | - |
dc.contributor.author | Wan, Qing | - |
dc.contributor.author | Lee, Sang-Heon | - |
dc.contributor.author | Choe, Yoonsuck | - |
dc.date.accessioned | 2023-12-26T18:42:50Z | - |
dc.date.available | 2023-12-26T18:42:50Z | - |
dc.date.created | 2022-01-07 | - |
dc.date.issued | 2021-12-08 | - |
dc.identifier.isbn | 9783030922726 | - |
dc.identifier.issn | 0302-9743 | - |
dc.identifier.uri | http://hdl.handle.net/20.500.11750/46883 | - |
dc.description.abstract | With the continuing improvement in deep learning methods in recent years, face recognition performance is starting to surpass human performance. However, current state-of-the-art approaches are usually trained on high-quality still images and do not work well in unconstrained video face recognition. We propose to use audio information in the video to aid in the face recognition task with mixed quality inputs. We introduce an Audio-Visual Aggregation Network (AVAN) to aggregate multiple facial and voice information to improve face recognition performance. To effectively train and evaluate our approach, we constructed an Audio-Visual Face Recognition dataset. Empirical results show that our approach significantly improves the face recognition accuracy on unconstrained videos. © 2021, Springer Nature Switzerland AG. | - |
dc.language | English | - |
dc.publisher | Springer Science and Business Media Deutschland GmbH | - |
dc.title | Video Face Recognition with Audio-Visual Aggregation Network | - |
dc.type | Conference Paper | - |
dc.identifier.doi | 10.1007/978-3-030-92273-3_13 | - |
dc.identifier.scopusid | 2-s2.0-85121919369 | - |
dc.identifier.bibliographicCitation | 28th International Conference on Neural Information Processing, ICONIP 2021, pp.150 - 161 | - |
dc.identifier.url | https://iconip2021.apnns.org/ | - |
dc.citation.conferencePlace | IO | - |
dc.citation.conferencePlace | Bali | - |
dc.citation.endPage | 161 | - |
dc.citation.startPage | 150 | - |
dc.citation.title | 28th International Conference on Neural Information Processing, ICONIP 2021 | - |
There are no files associated with this item.