VanFace

truckspy · Feb 18, 2021 · bb5b1be · bb5b1be
1 parent 98abdbc
commit bb5b1be
Show file tree

Hide file tree

Showing 5 changed files with 350 additions and 0 deletions.
diff --git a/detect_facemarks_by_dnn_vanface.php b/detect_facemarks_by_dnn_vanface.php
@@ -0,0 +1,59 @@
+<?php
+
+use CV\Scalar, CV\Size;
+use function CV\{imread, imwrite, cvtColor, circle};
+
+$netDet = \CV\DNN\readNetFromCaffe('models/ssd/res10_300x300_ssd_deploy.prototxt', 'models/ssd/res10_300x300_ssd_iter_140000.caffemodel');
+$netRecogn = \CV\DNN\readNetFromCaffe('models/VanFace/VanFace.prototxt', 'models/VanFace/VanFace.caffemodel');
+
+$src = imread("images/faces.jpg");
+$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 - $startY));
+        \CV\resize($face, $resized, new Size(60,60));
+        $resized = cvtColor($resized, \CV\COLOR_BGR2GRAY, 2);
+        \CV\meanStdDev($resized, $mean, $sdv);
+
+        $m = $mean->data()[0];
+        $s = $sdv->data()[0];
+
+        $blob = \CV\DNN\blobFromImage($resized, 1 / (0.000001 + $s),  new Size(60,60), new Scalar($m, $m, $m));
+        $netRecogn->setInput($blob);
+        //var_export($blob);die();
+        $out = $netRecogn->forward();
+        $data = $out->data();
+
+        for ($j=0;$j<68;$j++) {
+            $point = new \CV\Point($startX + $data[$j*2] * $face->cols, $startY + $data[$j*2+1] * $face->rows);
+            circle($src, $point, 2, new Scalar(0, 0, 255), 2);
+            //var_export($point);
+        }
+    }
+}
+
+imwrite("results/_detect_facemarks_by_dnn_vanface.jpg", $src);
diff --git a/models/README.md b/models/README.md
@@ -10,3 +10,4 @@ Sources:
 * ssdlite_mobilenet_v2_coco - https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md
 * ssd_mobilenet_v2_coco - https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md
 * insightface - https://github.com/axinc-ai/ailia-models/tree/master/face_identification/insightface
+* VanFace - https://github.com/lsy17096535/face-landmark
diff --git a/models/VanFace/VanFace.caffemodel b/models/VanFace/VanFace.caffemodel
diff --git a/models/VanFace/VanFace.prototxt b/models/VanFace/VanFace.prototxt
@@ -0,0 +1,290 @@
+name: "landmark"
+input: "data"
+input_dim: 1
+input_dim: 1
+input_dim: 60
+input_dim: 60
+########################################
+# the actual net
+# layer 1
+layer {
+    name: "Conv1"
+    type: "Convolution"
+    bottom: "data"
+    top: "Conv1"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    convolution_param {
+        num_output: 20
+        pad: 2
+        kernel_size: 5
+        stride: 1
+        weight_filler {
+            type: "xavier"
+            std: 0.1
+        }
+        bias_filler {
+            type: "constant"
+            value: 0.2
+        }
+    }
+}
+
+layer {
+  name: "ActivationTangH1"
+  bottom: "Conv1"
+  top: "ActivationTangH1"
+  type: "TanH"
+}
+
+layer {
+  name: "ActivationAbs1"
+  bottom: "ActivationTangH1"
+  top: "Abs1"
+  type: "AbsVal"
+}
+
+layer {
+    name: "Pool1"
+    type: "Pooling"
+    bottom: "Abs1"
+    top: "Pool1"
+    pooling_param {
+        pool: MAX
+        kernel_size: 2
+        stride: 2
+    }
+}
+
+layer {
+    name: "Conv2"
+    type: "Convolution"
+    bottom: "Pool1"
+    top: "Conv2"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    convolution_param {
+        num_output: 48
+        pad: 2
+        kernel_size: 5
+        stride: 1
+        weight_filler {
+        type: "xavier"
+        std: 0.1
+    }
+    bias_filler {
+        type: "constant"
+        value: 0.2
+        }
+    }
+}
+
+layer {
+  name: "ActivationTangH2"
+  bottom: "Conv2"
+  top: "ActivationTangH2"
+  type: "TanH"
+}
+
+layer {
+  name: "ActivationAbs2"
+  bottom: "ActivationTangH2"
+  top: "Abs2"
+  type: "AbsVal"
+}
+
+
+layer {
+    name: "Pool2"
+    type: "Pooling"
+    bottom: "Abs2"
+    top: "Pool2"
+    pooling_param {
+        pool: MAX
+        kernel_size: 2
+        stride: 2
+    }
+}
+
+# layer 3
+layer {
+    name: "Conv3"
+    type: "Convolution"
+    bottom: "Pool2"
+    top: "Conv3"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    convolution_param {
+        num_output: 64
+        pad: 0
+        kernel_size: 3
+        stride: 1
+        weight_filler {
+            type: "xavier"
+            std: 0.1
+        }
+        bias_filler {
+            type: "constant"
+            value: 0.2
+        }
+    }
+}
+
+
+layer {
+  name: "ActivationTangH3"
+  bottom: "Conv3"
+  top: "ActivationTangH3"
+  type: "TanH"
+}
+
+layer {
+  name: "ActivationAbs3"
+  bottom: "ActivationTangH3"
+  top: "Abs3"
+  type: "AbsVal"
+}
+
+layer {
+    name: "Pool3"
+    type: "Pooling"
+    bottom: "Abs3"
+    top: "Pool3"
+    pooling_param {
+        pool: MAX
+        kernel_size: 3
+        stride: 2
+    }
+}
+
+# layer 4
+layer {
+    name: "Conv4"
+    type: "Convolution"
+    bottom: "Pool3"
+    top: "Conv4"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    convolution_param {
+        num_output: 80
+        pad: 0
+        kernel_size: 3
+        stride: 1
+        weight_filler {
+            type: "xavier"
+            std: 0.1
+        }
+        bias_filler {
+            type: "constant"
+            value: 0.2
+        }
+    }
+}
+
+
+layer {
+  name: "ActivationTangH4"
+  bottom: "Conv4"
+  top: "ActivationTangH4"
+  type: "TanH"
+}
+
+layer {
+  name: "ActivationAbs4"
+  bottom: "ActivationTangH4"
+  top: "Abs4"
+  type: "AbsVal"
+}
+
+
+########################################
+
+layer {
+    name: "Dense1"
+    type: "InnerProduct"
+    bottom: "Abs4"
+    top: "Dense1"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    inner_product_param {
+        num_output: 512
+        weight_filler {
+            type: "xavier"
+        }
+        bias_filler {
+            type: "constant"
+            value: 0
+        }
+    }
+}
+
+
+layer {
+  name: "ActivationTangH5"
+  bottom: "Dense1"
+  top: "ActivationTangH5"
+  type: "TanH"
+}
+
+layer {
+  name: "ActivationAbs5"
+  bottom: "ActivationTangH5"
+  top: "Abs5"
+  type: "AbsVal"
+}
+
+
+layer {
+    name: "Dense3"
+    type: "InnerProduct"
+    bottom: "Abs5"
+    top: "Dense3"
+    param {
+        lr_mult: 1
+        decay_mult: 1
+    }
+    param {
+        lr_mult: 2
+        decay_mult: 0
+    }
+    inner_product_param {
+        num_output: 136
+        weight_filler {
+            type: "xavier"
+        }
+        bias_filler {
+            type: "constant"
+            value: 0
+        }
+    }
+}
diff --git a/results/detect_facemarks_by_dnn_vanface.jpg b/results/detect_facemarks_by_dnn_vanface.jpg