Name
Abstract | ||
|---|---|---|
Learning from demonstration (LfD) is useful in settings where hand-coding behaviour or a reward function is impractical. It has succeeded in a wide range of problems but typically relies on manually generated demonstrations or specially deployed sensors and has not generally been able to leverage the copious demonstrations available in the wild: those that capture behaviours that were occurring anyway using sensors that were already deployed for another purpose, e.g., traffic camera footage capturing demonstrations of natural behaviour of vehicles, cyclists, and pedestrians. We propose video to behaviour (ViBe), a new approach to learn models of behaviour from unlabelled raw video data of a traffic scene collected from a single, monocular, initially uncalibrated camera with ordinary resolution. Our approach calibrates the camera, detects relevant objects, tracks them through time, and uses the resulting trajectories to perform LfD, yielding models of naturalistic behaviour. We apply ViBe to raw videos of a traffic intersection and show that it can learn purely from videos, without additional expert knowledge. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1109/ICRA.2019.8794412 | 2019 International Conference on Robotics and Automation (ICRA) |
Keywords | Field | DocType |
uncalibrated camera,learning from demonstration,ViBe,traffic intersection,knowledge expert,video to behaviour,natural behaviour,reward function,hand-coding behaviour,wild,raw videos,naturalistic behaviour,LfD,monocular camera,single camera,traffic scene | Computer vision,Traffic camera,Learning from demonstration,Artificial intelligence,Traffic scene,Monocular,Machine learning,Mathematics | Journal |
Volume | Issue | ISSN |
abs/1811.03516 | 1 | 1050-4729 |
ISBN | Citations | PageRank |
978-1-5386-8176-3 | 3 | 0.73 |
References | Authors | |
32 | 11 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Feryal Behbahani | 1 | 3 | 1.07 |
| Kyriacos Shiarlis | 2 | 26 | 3.90 |
| Xi Chen | 3 | 86 | 10.34 |
| Vitaly Kurin | 4 | 13 | 4.95 |
| Sudhanshu Kasewa | 5 | 3 | 0.73 |
| Ciprian Stirbu | 6 | 3 | 0.73 |
| João Gomes | 7 | 16 | 6.24 |
| Supratik Paul | 8 | 8 | 3.52 |
| Frans A. Oliehoek | 9 | 397 | 40.32 |
| João V. Messias | 10 | 26 | 4.77 |
| Shimon Whiteson | 11 | 1460 | 99.00 |