Face detection has been a core problem in computer vision for more than a decade.
Not only has there been substantial progress in research, but many techniques for
face detection have also made their way into commercial products. Despite this
maturity, the algorithms for face detection remain difficult to
compare, and somewhat brittle to the specific conditions under which they are
applied. One difficulty in comparing different face detection algorithms is the lack
of enough detail to reproduce the published results, which makes it important to
establish better benchmarks of performance.
In this workshop, we introduce a new, challenging data set of images with faces
in unconstrained settings. A rigorous evaluation of different face detection
algorithms on this benchmark will emphasize the two main objectives of this workshop:
(1) establish the current state-of-the-art in face detection, and
(2) identify new frontiers of research in face detection.
To encourage an easy access of these face detection systems to the research community,
this workshop will present a cash award for best performing face detector
(please visit the FDDB page
for further details).
Also, there will be a best paper award. Both of these
awards are sponsored by
Microsoft Research India.