Resumen
Motivated by the need for photo-realistic simulation in autonomous driving, in this paper we present a video inpainting algorithm AutoRemover, designed specifically for generating street-view videos without any moving objects. In our setup we have two challenges: the first is the shadow, shadows are usually unlabeled but tightly coupled with the moving objects. The second is the large ego-motion in the videos. To deal with shadows, we build up an autonomous driving shadow dataset and design a deep neural network to detect shadows automatically. To deal with large ego-motion, we take advantage of the multi-source data, in particular the 3D data, in autonomous driving. More specifically, the geometric relationship between frames is incorporated into an inpainting deep neural network to produce high-quality structurally consistent video output. Experiments show that our method outperforms other state-of-the-art (SOTA) object removal algorithms, reducing the RMSE by over 19%.
| Idioma original | English |
|---|---|
| Título de la publicación alojada | AAAI 2020 - 34th AAAI Conference on Artificial Intelligence |
| Páginas | 12853-12861 |
| Número de páginas | 9 |
| ISBN (versión digital) | 9781577358350 |
| Estado | Published - 2020 |
| Evento | 34th AAAI Conference on Artificial Intelligence, AAAI 2020 - New York, United States Duración: feb 7 2020 → feb 12 2020 |
Serie de la publicación
| Nombre | AAAI 2020 - 34th AAAI Conference on Artificial Intelligence |
|---|
Conference
| Conference | 34th AAAI Conference on Artificial Intelligence, AAAI 2020 |
|---|---|
| País/Territorio | United States |
| Ciudad | New York |
| Período | 2/7/20 → 2/12/20 |
Nota bibliográfica
Publisher Copyright:Copyright c 2020, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
Financiación
Weiwei Xu is partially supported by NSFC (No. 61732016) and the fundamental research fund for the central universities. Jinhui Yu is partially supported by NSFC (No. 61772463). We also thank the reviewers and all the people who offered help.
| Financiadores | Número del financiador |
|---|---|
| National Natural Science Foundation of China (NSFC) | 61732016 |
| Fundamental Research Funds for the Central Universities | 61772463 |
ASJC Scopus subject areas
- Artificial Intelligence
Huella
Profundice en los temas de investigación de 'AutoRemover: Automatic object removal for autonomous driving videos'. En conjunto forman una huella única.Citar esto
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver