TY  - JOUR
AU  - David Vazquez
AU  - Javier Marin
AU  - Antonio Lopez
AU  - Daniel Ponsa
AU  - David Geronimo
PY  - 2014//
TI  - Virtual and Real World Adaptation for Pedestrian Detection
T2  - TPAMI
JO  - IEEE Transactions on Pattern Analysis and Machine Intelligence
SP  - 797
EP  - 809
VL  - 36
IS  - 4
KW  - Domain Adaptation
KW  - Pedestrian Detection
N2  - Pedestrian detection is of paramount interest for many applications. Most promising detectors rely on discriminatively learnt classifiers, i.e., trained with annotated samples. However, the annotation step is a human intensive and subjective task worth to be minimized. By using virtual worlds we can automatically obtain precise and rich annotations. Thus, we face the question: can a pedestrian appearance model learnt in realistic virtual worlds work successfully for pedestrian detection in realworld images?. Conducted experiments show that virtual-world based training can provide excellent testing accuracy in real world, but it can also suffer the dataset shift problem as real-world based training does. Accordingly, we have designed a domain adaptation framework, V-AYLA, in which we have tested different techniques to collect a few pedestrian samples from the target domain (real world) and combine them with the many examples of the source domain (virtual world) in order to train a domain adapted pedestrian classifier that will operate in the target domain. V-AYLA reports the same detection accuracy than when training with many human-provided pedestrian annotations and testing with real-world images of the same domain. To the best of our knowledge, this is the first work demonstrating adaptation of virtual and real worlds for developing an object detector.
SN  - 0162-8828
L1  - http://refbase.cvc.uab.es/files/vml2013.pdf
UR  - http://dx.doi.org/10.1109/TPAMI.2013.163
N1  - ADAS; 600.057; 600.054; 600.076
ID  - David Vazquez2014
ER  -