ZF/OpenFace
1
0
Fork 0
mirror of https://gitcode.com/gh_mirrors/ope/OpenFace.git synced 2026-05-16 12:17:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

Adding bounding box correction and an addition to non maximum supression.

Browse Source
This commit is contained in:
Tadas Baltrusaitis
2017-08-15 11:44:05 +01:00
parent dd17680e2c
commit 0c2137d950
1 changed files with 24 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
lib/local/LandmarkDetector/src/FaceDetectorMTCNN.cpp
View File

@@ -570,7 +570,7 @@ void FaceDetectorMTCNN::Read(string location)
}
// Perform non maximum supression on proposal bounding boxes prioritizing boxes with high score/confidence
std::vector<int> non_maximum_supression(const std::vector<cv::Rect_<float> >& original_bb, const std::vector<float>& scores, float thresh)
std::vector<int> non_maximum_supression(const std::vector<cv::Rect_<float> >& original_bb, const std::vector<float>& scores, float thresh, bool minimum)
{
// Sort the input bounding boxes by the detection score, using the nice trick of multimap always being sorted internally
@@ -600,7 +600,15 @@ std::vector<int> non_maximum_supression(const std::vector<cv::Rect_<float> >& or
const cv::Rect& rect2 = original_bb[pos->second];
float intArea = (rect1 & rect2).area();
float unionArea = rect1.area() + rect2.area() - intArea;
float unionArea;
if (minimum)
{
unionArea = cv::min(rect1.area(), rect2.area());
}
else
{
unionArea = rect1.area() + rect2.area() - intArea;
}
float overlap = intArea / unionArea;
// Remove the bounding boxes with less confidence but with significant overlap with the current one
@@ -787,7 +795,7 @@ bool FaceDetectorMTCNN::DetectFaces(vector<cv::Rect_<double> >& o_regions, const
generate_bounding_boxes(proposal_boxes, scores, proposal_corrections, prob_heatmap, corrections_heatmap, scale, t1, face_support);
// Perform non-maximum supression on proposals in this scale
vector<int> to_keep = non_maximum_supression(proposal_boxes, scores, 0.5);
vector<int> to_keep = non_maximum_supression(proposal_boxes, scores, 0.5, false);
select_subset(to_keep, proposal_boxes, scores, proposal_corrections);
proposal_boxes_all.insert(proposal_boxes_all.end(), proposal_boxes.begin(), proposal_boxes.end());
@@ -799,7 +807,7 @@ bool FaceDetectorMTCNN::DetectFaces(vector<cv::Rect_<double> >& o_regions, const
// Preparation for RNet step
// Non maximum supression accross bounding boxes, and their offset correction
vector<int> to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7);
vector<int> to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7, false);
select_subset(to_keep, proposal_boxes_all, scores_all, proposal_corrections_all);
apply_correction(proposal_boxes_all, proposal_corrections_all, false);
@@ -855,7 +863,7 @@ bool FaceDetectorMTCNN::DetectFaces(vector<cv::Rect_<double> >& o_regions, const
select_subset(to_keep, proposal_boxes_all, scores_all, proposal_corrections_all);
// Non maximum supression accross bounding boxes, and their offset correction
to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7);
to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7, false);
select_subset(to_keep, proposal_boxes_all, scores_all, proposal_corrections_all);
apply_correction(proposal_boxes_all, proposal_corrections_all, false);
@@ -912,12 +920,20 @@ bool FaceDetectorMTCNN::DetectFaces(vector<cv::Rect_<double> >& o_regions, const
apply_correction(proposal_boxes_all, proposal_corrections_all, true);
// Non maximum supression accross bounding boxes, and their offset correction
to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7);
to_keep = non_maximum_supression(proposal_boxes_all, scores_all, 0.7, true);
select_subset(to_keep, proposal_boxes_all, scores_all, proposal_corrections_all);
// TODO diff selection criteria for supression
// Correct the box to expectation to be tight around facial landmarks
for (size_t k = 0; k < proposal_boxes_all.size(); ++k)
{
proposal_boxes_all[k].x = proposal_boxes_all[k].width * -0.0075 + proposal_boxes_all[k].x;
proposal_boxes_all[k].y = proposal_boxes_all[k].height * 0.2459 + proposal_boxes_all[k].y;
proposal_boxes_all[k].width = 1.0323 * proposal_boxes_all[k].width;
proposal_boxes_all[k].height = 0.7751 * proposal_boxes_all[k].height;
// TODO correct the box to expectation
o_regions.push_back(cv::Rect_<double>(proposal_boxes_all[k].x, proposal_boxes_all[k].y, proposal_boxes_all[k].width, proposal_boxes_all[k].height));
o_confidences.push_back(scores_all[k]);
}
return true;