2012 Second International Conference on 3D Imaging, Modeling, Processing, Visualization &Amp; Transmission 2012
DOI: 10.1109/3dimpvt.2012.23
|View full text |Cite
|
Sign up to set email alerts
|

Classification and Pose Estimation of Vehicles in Videos by 3D Modeling within Discrete-Continuous Optimization

Abstract: This paper presents a framework for classification and pose estimation of vehicles in videos by assuming their given 3D models. We rank possible poses and types for each frame and exploit temporal coherence between consecutive frames for refinement. As a novelty, first, we cast the estimation of a vehicle's pose and type as a solution of a continuous optimization problem over space and time. Due to a non-convexity of this problem, good initial starting points are important. We propose to obtain them by a discr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
17
0
1

Year Published

2013
2013
2021
2021

Publication Types

Select...
3
1
1

Relationship

2
3

Authors

Journals

citations
Cited by 9 publications
(18 citation statements)
references
References 19 publications
0
17
0
1
Order By: Relevance
“…As others [12,9], we assume, that given the ground plane, the vehicle's pose is parameterized by p = (x, y, α) as shown in Fig. 2.…”
Section: Vehicle Pose Estimation In Videosmentioning
confidence: 99%
See 4 more Smart Citations
“…As others [12,9], we assume, that given the ground plane, the vehicle's pose is parameterized by p = (x, y, α) as shown in Fig. 2.…”
Section: Vehicle Pose Estimation In Videosmentioning
confidence: 99%
“…1). Different to [9], the pose estimation does not rely on the correct 3D model being in the training set. We have a generic classifier trained on multiple models, whereas [9] needs to have the correct model available and evaluates all models in various poses to obtain the correct pose.…”
Section: Introductionmentioning
confidence: 99%
See 3 more Smart Citations