As you advised, I've tried CIDetectorMinFeatureSize. It works as expected. Dict of parameters seems correct way to set the parameters.

    opt = {'CIDetectorAccuracy': 'CIDetectorAccuracyHigh',
           'CIDetectorMinFeatureSize': 0.2}
    d = CIDetector.detectorOfType_context_options_('CIDetectorTypeFace', None, opt)
    rects = d.featuresInImage_(ci_img)

The face can be detected when the param is small (0.1), not be detected when the param is large (>0.3). (Somehow larger parameters kill pythonista to home screen.)

I cannot still find the way to read the set option from CIDetector. If you know which API to use, could you show me?

Thank you so much.

Thanks,

I have no idea to read back the options from CIDetecotor (or maybe faceCoreDetector in this class), could you tell me how can I read the option after setting?

And I'll try to set CIDetectorMinFeatureSize later, sounds good to check it!

My code is in progress, here: https://www.dropbox.com/s/4jdeohur37kzflk/MaskFace.py?dl=0

Thank you,

Hi,

I want to specify options to CIDetector. Which is the correct way? (or both incorrect?)

options = {'CIDetectorAccuracy': 'CIDetectorAccuracyHigh'}
d = CIDetector.detectorOfType_context_options_('CIDetectorTypeFace', None, options)

or do I need to write as:

options = NSDictionary.dictionaryWithDictionary_({'CIDetectorAccuracy': 'CIDetectorAccuracyHigh'})
d = CIDetector.detectorOfType_context_options_('CIDetectorTypeFace', None, options)

Actually both looks working fine, but I cannot know whether the options is actually effective. I cannot tell the detection accuracy is better or not, from the result.

I'm new to iOS and pythonista, I'm very grad to be advised by experts.

Thanks,