Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

FaceDetector stuck if multiple faces present #3

Open
Pasmikh opened this issue Dec 2, 2021 · 0 comments
Open

FaceDetector stuck if multiple faces present #3

Pasmikh opened this issue Dec 2, 2021 · 0 comments

Comments

@Pasmikh
Copy link

Pasmikh commented Dec 2, 2021

Thank you for this repository!
Face Detector works fine when there is one face present, but when there are 2 or more Non maximum supression algorithm gets stuck in infinite while loop. It doesn't filter out non-intersecting boxes properly.
I have slightly modified your code in non_maximum_supression.dart to

var picks = <int>[];
  while (I.isNotEmpty) {
    var ind0 = positions.sublist(positions.length - 1, positions.length);
    var ind1 = positions.sublist(0, positions.length - 1);

    var xx1 = _maximum(_itemIndex(_x1, ind0)[0], _itemIndex(_x1, ind1));
    var yy1 = _maximum(_itemIndex(_y1, ind0)[0], _itemIndex(_y1, ind1));
    var xx2 = _minimum(_itemIndex(_x2, ind0)[0], _itemIndex(_x2, ind1));
    var yy2 = _minimum(_itemIndex(_y2, ind0)[0], _itemIndex(_y2, ind1));
    var w = _maximum(0.0, xx2 - xx1);
    var h = _maximum(0.0, yy2 - yy1);
    var inter = w * h;
    var o = inter /
        (_sum(_itemIndex(area, ind0)[0], _itemIndex(area, ind1)) - inter);

    picks.add(ind0[0]);

    List<double> newI = [];
    List<int> newPositions = [];
    o.asMap().forEach((key, value) {
      if (value <= threshold) {
        newI.add(I[key]);
        newPositions.add(positions[key]);
      }
    });

    I = newI;
    positions = newPositions;
  }

  List<Detection> results = [];

  for (var pick in picks) {
    results.add(detections[pick]);
  }
  return results;

and also in face_detection_service.dart


    for (var detection in detections) {
      Rect? bbox;
      final score = detection.score;
      if (score > threshold) {
        bbox = Rect.fromLTRB(
          inputImage.width * detection.xMin,
          inputImage.height * detection.yMin,
          inputImage.width * detection.width,
          inputImage.height * detection.height,
        );

        bbox = _imageProcessor.inverseTransformRect(
            bbox, image.height, image.width);
      }
      rectFaces.add({'bbox': bbox, 'score': score});
    }
    rectFaces.sort((a, b) => b['score'].compareTo(a['score']));

    return rectFaces;

It detects and returns multiple boxes for each face present. I hope it helps, thanks again for this project.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant