| Item type |
SIG Technical Reports(1) |
| 公開日 |
2022-09-29 |
| タイトル |
|
|
タイトル |
ニューラル場表現と表情類似度に基づく4次元ポートレート生成法 |
| タイトル |
|
|
言語 |
en |
|
タイトル |
4D portrait generation based on neural radiance field and facial expression similarity |
| 言語 |
|
|
言語 |
jpn |
| キーワード |
|
|
主題Scheme |
Other |
|
主題 |
計測・認証2 |
| 資源タイプ |
|
|
資源タイプ識別子 |
http://purl.org/coar/resource_type/c_18gh |
|
資源タイプ |
technical report |
| 著者所属 |
|
|
|
筑波大学システム情報工学研究群 |
| 著者所属 |
|
|
|
筑波大学計算科学研究センター |
| 著者所属 |
|
|
|
筑波大学計算科学研究センター |
| 著者所属(英) |
|
|
|
en |
|
|
Degree programs in Systems and Information Engineering, University of Tsukuba |
| 著者所属(英) |
|
|
|
en |
|
|
Center for Computational Sciences, University of Tsukuba |
| 著者所属(英) |
|
|
|
en |
|
|
Center for Computational Sciences, University of Tsukuba |
| 著者名 |
木村, 文哉
宍戸, 英彦
北原, 格
|
| 著者名(英) |
Fumiya, Kimura
Hidehiko, Shishido
Itaru, Kitahara
|
| 論文抄録 |
|
|
内容記述タイプ |
Other |
|
内容記述 |
This research aims to generate 4D portrait of a person that can be playback over a long time period. 4D portrait is a free-viewpoint video of a person with temporal changes in facial expression. In our proposed method, the parameters that represent facial expressions and head poses of a person are obtained from a video captured by a monocular RGB camera with continuously changing the viewpoint. Then, a neural radiance field (NeRF) is trained from the captured video and estimated parameters. Using this radiance field, 4D portrait is generated based on the similarity of the person's facial expressions. This paper describes a method for determining transition frames based on facial expression similarity to enable long playback. |
| 論文抄録(英) |
|
|
内容記述タイプ |
Other |
|
内容記述 |
This research aims to generate 4D portrait of a person that can be playback over a long time period. 4D portrait is a free-viewpoint video of a person with temporal changes in facial expression. In our proposed method, the parameters that represent facial expressions and head poses of a person are obtained from a video captured by a monocular RGB camera with continuously changing the viewpoint. Then, a neural radiance field (NeRF) is trained from the captured video and estimated parameters. Using this radiance field, 4D portrait is generated based on the similarity of the person's facial expressions. This paper describes a method for determining transition frames based on facial expression similarity to enable long playback. |
| 書誌レコードID |
|
|
収録物識別子タイプ |
NCID |
|
収録物識別子 |
AA12049625 |
| 書誌情報 |
研究報告エンタテインメントコンピューティング（EC）
巻 2022-EC-65,
号 38,
p. 1-6,
発行日 2022-09-29
|
| ISSN |
|
|
収録物識別子タイプ |
ISSN |
|
収録物識別子 |
2188-8914 |
| Notice |
|
|
|
SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc. |
| 出版者 |
|
|
言語 |
ja |
|
出版者 |
情報処理学会 |
IPSJ-EC22065038.pdf (2.3 MB)
