Fusion of Motion and Appearance for Robust People Detection in Cluttered Scenes - Intelligent Video Event Analysis and Understanding - page 123

ROC curves for pedestrian detection

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

Our detector
Zhang&Gong (vs2006)
Dalal & Triggs (CVPR2005)
Motion detector

0.1

0

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

False positive rate

−5

x 10

Fig. 11 ROC curves on the sequences of the underground ticket office for both the dynamic
and static people detectors as well as a pure motion based people detector.

spatial pyramid Gaussian Mixture approach to model variations in long-term motion
of human movement, which is computed via an improved background modeling.
Our model utilises both human appearance and their long-term motion information
through a fusion formulated in a Bayesian framework. Experiments demonstrate
that our method reduces significantly the false positive rate compared to that of
the state of the art static human detector under very challenging conditions. Note
that our current people appearance is modeled by histograms of oriented gradients,
which is an extension of the static detector proposed by Dalal and Triggs. However,
in principle, any other type of static detectors can be used in our framework. As our
model is based on long-term motion information therefore it requires a fixed camera
view during detection. Building a hybrid model of both long-term and short-term
motion information could possibly give more robust detections and also be adaptive
to some background and viewpoint change.

References

1. Boiman, O., Irani, M.: Detecting irregularities in images and in video. In: International
Conference on Computer Vision, pp. 462-469 (2005)
2. Cutler, R., Davis, L.: Robust real-time periodic motion detection: Analysis and applica-
tions. IEEE Transactions on Pattern Analysis and Machine Intelligence 22(8), 781-796
(2000)
3. Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: Inter-
national Conference on Computer Vision & Pattern Recognition, vol. 2, pp. 886-893
(2005)
4. Dalal, N., Triggs, B., Schmid, C.: Human detection using oriented histograms of flow and
appearance. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006. LNCS, vol. 3952,
pp. 428-441. Springer, Heidelberg (2006)