Few-Shots Novel Space-Time View Synthesis from Consecutive Photos
Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 240-249.
Sparad:
| Huvudupphovsmän: | , |
|---|---|
| Materialtyp: | Bài viết |
| Språk: | English |
| Publicerad: |
Springer Nature
2023
|
| Ämnen: | |
| Länkar: | https://link.springer.com/chapter/10.1007/978-3-031-36886-8_20 http://elib.vku.udn.vn/handle/123456789/2729 |
| Taggar: |
Lägg till en tagg
Inga taggar, Lägg till första taggen!
|
| Thư viện lưu trữ: | Trường Đại học Công nghệ Thông tin và Truyền thông Việt Hàn - Đại học Đà Nẵng |
|---|
| id |
oai:elib.vku.udn.vn:123456789-2729 |
|---|---|
| record_format |
dspace |
| spelling |
oai:elib.vku.udn.vn:123456789-27292023-09-26T01:45:43Z Few-Shots Novel Space-Time View Synthesis from Consecutive Photos Mai, Van Quan Nguyen, Duc Dung NeRF View synthesis Few-shot view reconstruction Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 240-249. Despite the remarkable result of Neural Scene Flow Fields [10] in novel space-time view synthesis of dynamic scenes, the model has limited ability when a few input views are provided. To enable the few-shots novel space-time view synthesis of dynamic scenes, we propose a new approach that extends the model architecture to use shared priors learned across scenes to predict appearance and geometry at static background regions. Throughout the optimization, our network is trained to rely on the image features extracted from a few input views or from the learned knowledge for reconstructing unseen regions based on the camera view direction. We conduct multiple experiments on NVIDIA Dynamic Scenes Dataset [23] that demonstrate our approach results in a better rendering quality compared to the prior work when a few input views are available. 2023-09-26T01:45:36Z 2023-09-26T01:45:36Z 2023-07 Working Paper 978-3-031-36886-8 https://link.springer.com/chapter/10.1007/978-3-031-36886-8_20 http://elib.vku.udn.vn/handle/123456789/2729 en application/pdf Springer Nature |
| institution |
Trường Đại học Công nghệ Thông tin và Truyền thông Việt Hàn - Đại học Đà Nẵng |
| collection |
DSpace |
| language |
English |
| topic |
NeRF View synthesis Few-shot view reconstruction |
| spellingShingle |
NeRF View synthesis Few-shot view reconstruction Mai, Van Quan Nguyen, Duc Dung Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| description |
Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 240-249. |
| format |
Working Paper |
| author |
Mai, Van Quan Nguyen, Duc Dung |
| author_facet |
Mai, Van Quan Nguyen, Duc Dung |
| author_sort |
Mai, Van Quan |
| title |
Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| title_short |
Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| title_full |
Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| title_fullStr |
Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| title_full_unstemmed |
Few-Shots Novel Space-Time View Synthesis from Consecutive Photos |
| title_sort |
few-shots novel space-time view synthesis from consecutive photos |
| publisher |
Springer Nature |
| publishDate |
2023 |
| url |
https://link.springer.com/chapter/10.1007/978-3-031-36886-8_20 http://elib.vku.udn.vn/handle/123456789/2729 |
| _version_ |
1849196497506140160 |