Vojtech Franc presents Learning CNNs from Weakly Annotated Facial Images

On 2018-12-18 11:00 at G205, Karlovo náměstí 13, Praha 2
Learning of convolutional neural networks (CNNs) to perform a face
recognition task requires a large set of facial images each annotated with
a label to be predicted. In this talk we describe a method for learning
CNNs from weakly annotated images. The weak annotation in our setting means
that a pair of an attribute label and a person identity label is assigned
to a set of faces automatically detected in the image. The challenge is to
link the annotation with the correct face. The weakly annotated images of
this type can be collected by an automated process not requiring human
labour. We formulate learning from weakly annotated images as maximum
likelihood (ML) estimation of a parametric distribution describing the
weakly annotated images. The ML problem can be solved by an instance of the
EM algorithm which in its inner loop learns a CNN to predict attribute
labels from facial images. Experiments on age and gender estimation problem
show that the proposed algorithm significantly outperforms the existing
heuristic approach for dealing with this type of data. A practical outcome
of our paper is a new annotation of the IMDB database containing 300k faces
each one annotated by biological age, gender and identity labels.
Za obsah zodpovídá: Petr Pošík