An image may contain several faces captioned
with their corresponding names. It may so happen that a
facial image may be wrongly annotated. The self regulated
image face naming technique that we propose aims at
labeling a face in the image accurately. This is a challenging
task because of the very large appearance variation in the
images, as well as the potential mismatch between images
and their captions. We propose this efficient face naming
technique which is self regulated and aims at correctly
labeling a face in an image. We first propose a new method
called Unsupervised Regularized Low-Rank Depiction
(URLRD) which productively employs the wrongly named
image information to determine a low-rank matrix which is
obtained by recreation along with examining many subspace
structures of the data. Certain circumstances befall where a
face is recreated by using its own facial image or from other
subject’s facial images. From the recreation method used we
deduce a discriminatory matrix. Besides this we also deploy
the Large Margin Nearest Neighbor (LMNN) method for
face labeling an image which further leads to yet another
kernel matrix and is based on the Mahalanobis distances of
the data. We can note that the two corresponding facial
matrices can be combined in such a way as to enhance the
quality of each other. The fused matrix is used as a new
reiterative plan to deduce the names of each facial image.
Extensive analysis demonstrates the effectiveness of our
accession.
Kavitha G L : is an MTech (CSE) graduate, pursuing PhD in cloud
computing. She is currently working as Asst. Professor in Atria
Institute of Technology, Bangalore. Her area of interests includes
machine learning and image processing.
D V Pranathi Suhasini : is pursuing B.E in Atria Institute of
Technology, Bangalore. Her area of interests includes image
processing.
Seema B Nikam : is pursuing B.E in Atria Institute of Technology,
Bangalore. Her area of interests includes image processing.
In this paper, we present an approach for face detection
and naming which minimizes computation time while
achieving high detection accuracy. To productively
employ the face naming of the facial images we introduce URLRD by using this scheme we increase the evaluation
of auto face annotation performance. We also intensify the
LMNN algorithm which delves on discriminating
Mahalanobis distance metric. Two facial matrices are
obtained by merging the matrices acquired by URLRD
and LMNN. Our proposed methods focus on tackling the
critical problem of enhancing the label quality and
accurately naming the facial images. We analyze the two
challenging and interesting real-world datasets from
which we can certify that our URLRD and LMNN
outperforms ULR and kNN respectively and several other
baseline algorithms. Our future work will further speed up
the current solution for very large applications and
investigate other techniques to improve the label
refinement task.
[1] P. Viola and M. J. Jones, “Robust real-time face
detection,” Int. J. Comput. Vis., vol. 57, no. 2, pp. 137–
154, 2004
[2] G. Liu, Z. Lin, and Y. Yu, “Robust subspace
segmentation by low-rank representation,” in Proc. 27th
Int. Conf. Mach. Learn., Haifa, Israel, Jun. 2010, pp.
663–670.
[3] Dayong WANG, Chu Hong HOI, Ying HE, Jianke
ZHU “Mining Weakly-Labeled Web Facial Images
for Search-Based Face Annotation”,2014.
[4] K. Q. Weinberger and L. K. Saul, “Distance metric
learning for large margin nearest neighbor
classification,” J. Mach. Learn. Res., vol. 10, pp. 207–
244, Feb. 2009.
[5] T. L. Berget al., “Names and faces in the news,” in Proc.
17th IEEE Comput. Soc. Conf. Comput. Vis. Pattern
Recognit., Washington, DC, USA, Jun./Jul. 2004, pp. II-
848–II-854.
[6] M. Guillaumin, J. Verbeek, and C. Schmid, “Multiple
instance metric learning from automatically labeled bags
of faces,” in Proc. 11th Eur. Conf. Comput. Vis.,
Heraklion, Crete, Sep. 2010, pp. 634–647.
[7] D. Ozkan and P. Duygulu, “A graph based approach for
naming faces in news photos,” in Proc. 19th IEEE
Comput. Soc. Conf. Comput. Vis. Pattern Recognit.,
New York, NY, USA, Jun. 2006, pp. 1477–1482
[8] Z. Zeng et al., “Learning by associating ambiguously
labeled images,” in Proc. 26th IEEE Conf. Comput. Vis.
Pattern Recognit., Portland, OR, USA, Jun. 2013, pp.
708–715
[9] J. Luo and F. Orabona, “Learning from candidate
labeling sets,” in Proc. 23rd Annu. Conf. Adv. Neural Inf.
Process. Syst., Vancouver, BC, Canada, Dec. 2010, pp.
1504–1512
[10] M. Guillaumin, T. Mensink, J. Verbeek, and C. Schmid,
“Face recognition from caption-based supervision,” Int. J.
Comput. Vis., vol. 96, no. 1, pp. 64–82, 2012
[11] M. Guillaumin, T. Mensink, J. Verbeek, and C. Schmid,
“Automatic face naming with caption-based
supervision,” in Proc. 21st IEEE Conf. Comput. Vis.
Pattern Recognit., Anchorage, AL, USA, Jun. 2008, pp.
1–8.
