Change Log
Contents
Change Log#
0.4.0#
Major version breaking release!#
This release includes numerous bug fixes, api updates, and code base changes make it largely incompatible with previous releases
To fork development from an older version of
py-feat
you can use this archival repo instead
New#
Added
animate_face
andplot_face
functions infeat.plotting
moduleFex
data-classes returned fromDetector.detect_image()
orDetector.detect_video()
now store the names of the different detectors used as attributes:.face_model
,.au_model
, etcThe AU visualization model used by
plot_face
andDetector.plot_detections(faces='aus')
has been updated to include AU11 and remove AU18 making it consistent with Py-feat’s custom AU detectors (svm
andlogistic
)A new AU visualization model supporting the
jaanet
AU detector, which only has 12 AUs, has now been added and will automatically be used ifDetector(au_model='jaanet')
.This visualization model can also be used by the
plot_face
function by by passing it to themodel
argument:plot_face(model='jaanet_aus_to_landmarks')
Breaking Changes#
Detector
no longer support unintialized models, e.g.any_model = None
This is is also true for
Detector.change_model
Columns of interest on
Fex
data classes were previously accessed like class methods, i.e.fex.aus()
. These have now been changed to class attributes, i.e.fex.aus
Remove support for
DRML
AU detectorRemove support for
RF
AU and emotion detectorsNew default detectors:
svm
for AUsresmasknet
for emotionsimg2pose
for head-pose
Development changes#
Revamped pre-trained detector handling in new
feat.pretrained
moduleMore tests including testing all detector combinations
Fixes#
0.3.7#
Fix import error due to missing init
0.3.6#
Trigger Zenodo release
0.2.0#
Testing pypi upload