Sie sind nicht angemeldet Anmelden | Anmelden
max planck institut
informatik
┬╗ Home ┬╗ Departments ┬╗

Disparity Statistics for Pedestrian Detection: Combining Appearance, Motion and Stereo

Disparity Statistics for Pedestrian Detection: Combining Appearance, Motion and Stereo

Stefan Walk, Konrad Schindler and Bernt Schiele
in 11th  European Conference on Computer Vision (ECCV 2010), September 5-11, 2010, Hersonissos, Heraktion, Crete, Greece

Abstract: Pedestrian detection is an important problem in computer 

vision due to its importance for applications such as visual surveillance, 

robotics, and automotive safety. This paper pushes the state-of-the-art 

of pedestrian detection in two ways. First, we propose a simple yet highly 

e´ČÇective novel feature based on binocular disparity, outperforming previ- 

ously proposed stereo features. Second, we show that the combination of 

di´ČÇerent classi´Čüers often improves performance even when classi´Čüers are 

based on the same feature or feature combination. These two extensions 

result in signi´Čücantly improved performance over the state-of-the-art on 

two challenging datasets.

Material

 

Paper

 

Poster

 

Results and Source Code

 
 

 

Contact

For further information and questions please contact Stefan WalkKonrad Schindler or Bernt Schiele.

 

 

alte website