text
stringlengths
0
43
which O
hinders O
AM S-MANP
’ O
s S-MATE
broader O
adoption O
to O
critical O
structural B-CONPRI
component E-CONPRI
manufacturing S-MANP
. O
Therefore O
, O
it O
is O
crucial O
to O
detect O
any O
process B-CONPRI
change/anomaly E-CONPRI
in O
a O
timely O
and O
accurate S-CHAR
manner O
for O
potential O
corrective O
operations O
. O
Real-time O
thermal B-FEAT
image E-FEAT
streams O
captured O
from O
AM B-MANP
processes E-MANP
are O
regarded O
as S-MATE
most O
informative O
signatures O
of O
the O
process S-CONPRI
stability O
. O
Existing O
state-of-the-art S-CONPRI
studies O
on O
thermal B-FEAT
image E-FEAT
streams O
focus O
merely O
on O
in B-CONPRI
situ E-CONPRI
sensing O
, O
feature B-ENAT
extraction E-ENAT
, O
and O
their O
relationship O
with O
process S-CONPRI
setup O
parameters S-CONPRI
and O
material B-CONPRI
properties E-CONPRI
. O
The O
objective O
of O
this O
paper O
is O
to O
develop O
a O
statistical B-CONPRI
process I-CONPRI