| Robust Real-time Object
Detection [P. Viola, M. Jones]
Presented by: Jakob Erikson
ABSTRACT: "Viola and Jones use a boosted combination
of weak features to achieve good false-positive and detection
rates using a large set of primitive but extremely cheap features.
By combining increasingly complex classifiers into a ``cascade'',
they achieve real-time object detection performance, approx
15 times faster than any previous algorithm while achieving
I'll introduce two concepts of major importance and one successful
application of these during the presentation:
1. The use of boosting for improving the performance of a
2. Cascaded classifiers for drastic performance improvements
3. An example of how these techniques can be applied to face
detection using an innovative set of extremely cheap features.
The first two parts are probably relevant to most of us and
the third is a new approach to an old and well known problem.
So if you aren't already familiar with these things, you really
want to show up on Monday! =)