86 7. PERSONALIZED CAPSULE WARDROBE CREATION
from the wardrobe I
. Formally, we have
G
I
j‚
G
D
1
jO
j
X
o
i
2O
cmp.o
i
/; (7.8)
where o
i
is the i -th outfit, and cmp./ refers to the outfit compatibility. To measure cmp.o
i
/, we
adopt the compatibility indicator in [31], where each outfit is treated as a sequence of items and
each item is regarded as a time step input of a bidirectional LSTM. In particular, cmp.o
i
/ can
be computed as follows:
8
ˆ
<
ˆ
:
cmp.o
i
/ D E
f
o
i
I
f
C E
b
.
o
i
I
b
/
;
E
f
o
i
I
f
D
1
N
P
N
tD1
log Pr
o
i;tC1
jo
i;1
; : : : ; o
i;t
I
f
;
E
b
.
o
i
I
b
/
D
1
N
P
0
tDN 1
log Pr
.
o
i;t
jo
i;N
; : : : ; o
i;tC1
I
b
/
;
(7.9)
where Pr.j/ stands for the conditional probability. E
f
.o
i
I
f
/ and E
b
.o
i
I
b
/ refer to the for-
ward and backward probability that the outfit o
i
would be a compatible one.
7.3.3 PCW CREATION
Based on the user modeling and garment modeling that enable us to comprehensively measure
the overall compatibility of a given wardrobe, we can now proceed to present our framework
for the automatic PCW creation. In particular, we cast the PCW creation as a combinatorial
optimization problem and propose a heuristic PCW creation method, which is summarized in
Algorithm 7.4. e underlying philosophy is to delete items from the original wardrobe that
can degrade the overall wardrobe compatibility and add candidate items that can improve the
compatibility.
Considering the practical situation, we first set the maximum and minimum numbers
of items for each category in a wardrobe. For simplicity, here we uniformly set that as N
max
and N
min
for all categories. At each iteration, we first check whether the number of items of
each category (i.e., N
c
) in a wardrobe has reached the pre-assigned maximum and minimum
number (i.e., N
max
and N
min
). If N
c
< N
min
(N
c
> N
max
), the algorithm would add (delete)
one item of the category c that maximizes (maximally hurts) the overall wardrobe compatibility
according to our wardrobe compatibility scoring model S./. Otherwise, if N
min
N
c
N
max
,
the algorithm would check if there is an existing (unsuitable or redundant) item deteriorating
the compatibility and removing which would boost the wardrobe compatibility. If yes, the item
will be deleted. In light of this, this operation will adaptively adjust the number of items of each
category, making the final PCW meeting the user’s preferences over different item categories.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset