![](bg9.png)
For the time being, we recommend using either
Fern T alone (for speed), or combining this with Fern
C (for somewhat better ability to avoid failures). If
we are to make further progress, we need a new Fern
which is independent of both.
5 CONCLUSION
This remains work in progress: it raises more ques-
tions than it answers. Haar classifiers followed by
”hand-crafted ferns” can usually find the nose tip, but
are there other, faster, more accurate or more reli-
able ways? Is the nose tip the best feature to detect,
or would other parts of the nose (bridge or nostrils)
be better? To what extent can the nose alone deter-
mine head pose? - occluded nostrils mean the user is
looking down, and prominent nostrils mean the user
is looking up, but can this variation be detected with
sufficient accuracy to be useful?
To some extent, our boxcar method combines the
advantages of Haar Features and Random Ferns, but
as a first attempt, it is unlikely to be optimal. A deeper
study would be welcome, but this would require a re-
turn to first principles and is work for the future.
The results in Section 4 demonstrate that (a) our
combined cascade fails less often than MCS (Cas-
trill
´
on et al., 2007), and (b) it does not insist on a
wrong interpretation of problematic faces such as Fig-
ure 4a. This is progress. Nevertheless, there is clearly
much more work still to be done before we have a
reliable nose-finder.
The target for on-the-spot retraining has not been
met, particularly for profile noses. While we are
waiting for faster portable computers to become
widespread, we may have to accept that on-the-spot
retraining is limited to frontal noses.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the project: “Set-
ting up of transdisciplinary research and knowledge
exchange (TRAKE) complex at the University of
Malta (ERDF.01.124)”, which is co-financed by the
European Union through the European Regional De-
velopment Fund 2014–2020.
REFERENCES
Cao, X., Wei, Y., Wen, F., and Sun, J. (2012). Face align-
ment by explicit shape regression. In CVPR, pages
2887–2894.
Castrill
´
on, M., D
´
eniz, O., Hern
´
andez, M., and Guerra, C.
(2007). Encara2: Real-time detection of multiple
faces at different resolutions in video streams. In Jour-
nal of Visual Communication and Image Representa-
tion Vol 18 No 2, pages 130–140.
Caunce, A., Cristinacce, D., Taylor, C., and Cootes, T. F.
(2009). Locating facial features and pose estimation
using a 3d shape model. In ISVC09.
Cristina, S. and Camilleri, K. P. (2018). Unobtrusive and
pervasive video-based eye-gaze tracking. In Image
and Vision Computing 74, pages 21–40.
Huang, G. B., Ramesh, M., Berg, T., and Learned-Miller,
E. (2007). Labeled faces in the wild: A database for
studying face recognition in unconstrained environ-
ments. Technical Report 07-49, University of Mas-
sachusetts, Amherst.
Ishikawa, T., Baker, S., Matthews, I., and Kanade, T.
(2004). Passive driver gaze tracking with active ap-
pearance models. In Proceedings of the 11th World
Congress on Intelligent Transportation Systems.
Lienhart, R. and Maydt, J. (2002). An extended set of haar-
like features for rapid object detection. In Proceed-
ings. 2002 International Conference on Image Pro-
cessing volume 1, pages I–900. IEEE.
Lu, F. and Chen, Y. (2016). Person-independent eye gaze
prediction from eye images using patch-based fea-
tures. In Neurocomputing (NC).
OpenCV (2015). Open Source Computer Vision Library.
Ozuysal, M., Calonder, M., Lepetit, V., and Fua, P. (2010).
Fast key-point recognition using random ferns. In
IEEE Transactions on Pattern Analysis and Machine
Intelligence,32(3), pages 448–461.
Rahmad, C., Andrie, R., Putra, D., Dharma, I., Darmono,
H., and Muhiqqin, I. (2020). Comparison of viola-
jones haar cascade classifier and histogram of oriented
gradients (hog) for face detection. In IOP Confer-
ence Series: Materials Science and Engineering. 732.
012038.
Sapienza, M. and Camilleri, K. P. (2014). Fasthpe: A recipe
for quick head pose estimation. Technical Report TR-
SCE-2014-01, University of Malta.
Viola, P. and Jones, M. (2004). Robust real-time face de-
tection. In International Journal of Computer Vision,
57(2), pages 137–154.
Weidenbacher, U., Layher, G., Bayerl, P., and Neumann,
H. (2006). Detection of head pose and gaze direction
for human-computer interaction. In Perception and
Interactive Technologies. PIT.
Zhang, Y. (2016). Eye tracking and gaze interface design
for pervasive displays. PhD thesis, University of Lan-
caster.
As Plain as the Nose on Your Face?
479