1.
Jürgen ’Juxi’ Leitner
S. Harding, P. Chandrashekhariah
M. Frank, G. Spina,
A. Förster, J. Triesch, J. Schmidhuber
idsia / usi / supsi, machine intelligence, fias
autonomous learning
of robust visual object
detection & identification
on a humanoid
#icdl/epirob 2012

2.
our iCub
setup is for
manipulation

3.
thanks to G. Metta and IIT for this picture
visual
perception

4.
the
challenge

5.
IDSIA’s three parts
Harding et al., GPTP 2012,
Leitner et al., IROS 2012 Leitner et al., BICA 2012
Leitner et al., ICDL 2012

6.
current
cv approaches

7.
detecting
objects
Harding et al., GPTP 2012

8.
our learning
approach

9.
INP INP INP + dilate min avg
cartesian
genetic
programming

10.
INP INP INP + dilate min avg
Func,on" 3"
Connec,on"1" "#2"
Connec,on"2"
A"real"number"
"#1"
4.3" cartesian
genetic
"""
"
programming

11.
icImage GreenTeaBoxDetector::runFilter() {
! icImage node0 = InputImages[6];
! icImage node1 = InputImages[1];
! icImage node2 = node0.absdiff(node1);
! icImage node5 = node2.SmoothBilateral(11);
! icImage node12 = InputImages[0];
! icImage node16 = node12.Sqrt();
! icImage node33 = node16.erode(6);
! icImage node34 = node33.log();
! icImage node36 = node34.min(node5);
! icImage node49 = node36.Normalize();
//cleanup
...
icImage out = node49.threshold(230.7218f);
! return out;
}
detection

12.
detect

13.
detect

14.
BUT:
supervised learning
approach

15.
FIAScollaboration
saliency
map
feature
segmentation

16.
pre
segmentation

17.
combined
approach

18.
MoBeE
framework Frank et al., ICINCO, 2012.

19.
conclusion
salient object detection
+
rough feature-based segmentation
=
automatic training set generation
for
cgp-based, robust filter-learning

20.
thanks
for listening
juxi@idsia.ch http://Juxi.net/projects http://robotics.idsia.ch
video at http://robotics.idsia.ch/im-clever/