face_to_vector_by_dnn_insightface.php

<?php

use CV\Scalar, CV\Size;
use function CV\{imread, imwrite};

$netDet = \CV\DNN\readNetFromCaffe('models/ssd/res10_300x300_ssd_deploy.prototxt', 'models/ssd/res10_300x300_ssd_iter_140000.caffemodel');
$netRecogn = \CV\DNN\readNetFromONNX('models/insightface/arcface_mobilefacenet.onnx');

$src = imread("images/faces.jpg");
$size = $src->size(); // 2000x500

$minSide = min($size->width, $size->height);
$divider = $minSide / 300;
\CV\resize($src, $resized, new Size($size->width / $divider, $size->height / $divider)); // 1200x300

//var_export($resized);

$blob = \CV\DNN\blobFromImage($resized, 1, new Size(), new Scalar(104, 177, 123), true, false);

$netDet->setInput($blob);

$r = $netDet->forward();

//var_export($r->shape);

$faces = [];
$scalar = new Scalar(0, 0, 255);
for ($i = 0; $i < $r->shape[2]; $i++) {
    $confidence = $r->atIdx([0,0,$i,2]);
    if ($confidence > 0.9) {
        var_export($confidence);echo "\n";
        $startX = $r->atIdx([0,0,$i,3]) * $src->cols;
        $startY = $r->atIdx([0,0,$i,4]) * $src->rows;
        $endX = $r->atIdx([0,0,$i,5]) * $src->cols;
        $endY = $r->atIdx([0,0,$i,6]) * $src->rows;

        $face = $src->getImageROI(new \CV\Rect($startX, $startY, $endX - $startX, $endY - $startY));
        //imwrite("results/_face.jpg", $face);

        $blob = \CV\DNN\blobFromImage($face, 1.0 / 255, new Size(112, 112), new Scalar(), true, false);
        $netRecogn->setInput($blob);
        $vec = $netRecogn->forward();
        \CV\normalize($r, $r);

        //$vec->print();
        var_export($vec->data());
    }
}