2

I want to use Vision API in android to detect the face and the landmarks over the face. I followed the Vision API sample :

https://github.com/googlesamples/android-vision/tree/master/visionSamples/photo-demo/

My issues are:

1) I cannot understand the details of this object while debugging:

FaceDetector detector = new FaceDetector.Builder(context)
    .setTrackingEnabled(false)
    .setLandmarkType(FaceDetector.ALL_LANDMARKS)
    .setProminentFaceOnly(true)
    .build();

image that shows the details of 'detector'

Cannot understand 'zzbbc','zzbbd'...etc

2)

Frame frame = new Frame.Builder().setBitmap(bitmap).build();
SparseArray<Face> faces = detector.detect(frame);`

Here the size of faces is returned as zero.

There is no exception thrown, I can see the image but the rectangle and dots cannot be seen.

Can anyone please help me out with this issue?

ArnonZ
  • 3,822
  • 4
  • 32
  • 42
Neha
  • 91
  • 4

1 Answers1

0

zzbbc, zzbbd, etc. are internal details of the implementation that aren't meant to be inspected. You don't need to know what these are to use the API.

In this case, no faces were detected. Note that the "prominentFaceOnly" setting will mean that the detector only looks a single large face (i.e., filling greater than a third of the screen width). If the faces in your photo are smaller than this, they will not be detected.

pm0733464
  • 2,862
  • 14
  • 16
  • Thanks a lot for the prompt reply. After debugging I found that detector.isOperational() returns false which means "Face detector dependencies are not yet available.". Can you tell me what are the dependencies that are needed for the same? – Neha Oct 08 '15 at 11:19