private static void Load(string directory, out IList <Matrix <RgbPixel> > images, out IList <uint> labels)
{
var imageList = new List <Matrix <RgbPixel> >();
var labelList = new List <uint>();
foreach (var file in Directory.EnumerateFiles(directory))
{
var name = Path.GetFileName(file);
var s = name.Split('_');
if (s.Length != 4 || !uint.TryParse(s[1], out var gender))
{
continue;
}
using (var tmp = Dlib.LoadImageAsMatrix <RgbPixel>(file))
{
var m = new Matrix <RgbPixel>(Size, Size);
Dlib.ResizeImage(tmp, m);
imageList.Add(m);
labelList.Add(gender);
}
}
images = imageList;
labels = labelList;
}
private static void Preprocess(string type, string input, FrontalFaceDetector faceDetector, ShapePredictor posePredictor, string output)
{
var imageCount = 0;
var r = new ulong[Size * Size];
var g = new ulong[Size * Size];
var b = new ulong[Size * Size];
var csv = ReadCsv(Path.Combine(input, $"{type}.csv"));
var outputDir = Path.Combine(output, type);
foreach (var kvp in csv)
{
using (var tmp = Dlib.LoadImageAsMatrix <RgbPixel>(Path.Combine(input, type, kvp.Key)))
{
var dets = faceDetector.Operator(tmp);
if (!dets.Any())
{
Console.WriteLine($"Warning: Failed to detect face from '{kvp}'");
continue;
}
// Get max size rectangle. It could be better face.
var det = dets.Select((val, idx) => new { V = val, I = idx }).Aggregate((max, working) => (max.V.Area > working.V.Area) ? max : working).V;
using (var ret = posePredictor.Detect(tmp, det))
using (var chip = Dlib.GetFaceChipDetails(ret, Size, 0d))
using (var faceChips = Dlib.ExtractImageChip <RgbPixel>(tmp, chip))
{
var dst = Path.Combine(outputDir, kvp.Key);
var dstDir = Path.GetDirectoryName(dst);
Directory.CreateDirectory(dstDir);
Dlib.SaveJpeg(faceChips, Path.Combine(outputDir, kvp.Key), 100);
var index = 0;
for (var row = 0; row < Size; row++)
{
for (var col = 0; col < Size; col++)
{
var rgb = faceChips[row, col];
r[index] += rgb.Red;
g[index] += rgb.Green;
b[index] += rgb.Blue;
index++;
}
}
}
imageCount++;
}
}
using (var mean = new Matrix <RgbPixel>(Size, Size))
{
var index = 0;
for (var row = 0; row < Size; row++)
{
for (var col = 0; col < Size; col++)
{
var red = (double)r[index] / imageCount;
var green = (double)g[index] / imageCount;
var blue = (double)b[index] / imageCount;
var newRed = (byte)Math.Round(red);
var newGreen = (byte)Math.Round(green);
var newBlue = (byte)Math.Round(blue);
mean[row, col] = new RgbPixel(newRed, newGreen, newBlue);
index++;
}
}
Dlib.SaveBmp(mean, Path.Combine(output, $"{type}.mean.bmp"));
}
}
private static void Load(string type, string directory, string meanImage, out IList <Matrix <RgbPixel> > images, out IList <uint> labels)
{
Matrix <RgbPixel> mean = null;
try
{
if (File.Exists(meanImage))
{
mean = Dlib.LoadImageAsMatrix <RgbPixel>(meanImage);
}
var csv = ReadCsv(Path.Combine(directory, $"{type}.csv"));
var imageList = new List <Matrix <RgbPixel> >();
var labelList = new List <uint>();
foreach (var kvp in csv)
{
var path = Path.Combine(directory, type, kvp.Key);
if (!File.Exists(path))
{
continue;
}
using (var tmp = Dlib.LoadImageAsMatrix <RgbPixel>(path))
{
if (mean != null)
{
using (var m = new Matrix <RgbPixel>(Size, Size))
{
Dlib.ResizeImage(tmp, m);
// ToDo: Support subtract operator on DlibDotNet
// var ret = m - mean;
var ret = new Matrix <RgbPixel>(Size, Size);
for (var row = 0; row < Size; row++)
{
for (var col = 0; col < Size; col++)
{
var left = m[row, col];
var right = mean[row, col];
var red = left.Red - right.Red;
var green = left.Green - right.Green;
var blue = left.Blue - right.Blue;
ret[row, col] = new RgbPixel((byte)red, (byte)green, (byte)blue);
}
}
imageList.Add(ret);
}
}
else
{
var m = new Matrix <RgbPixel>(Size, Size);
Dlib.ResizeImage(tmp, m);
imageList.Add(m);
}
labelList.Add(kvp.Value);
}
}
images = imageList;
labels = labelList;
}
finally
{
mean?.Dispose();
}
}
static void Main(string[] args)
{
/// FaceDetectionWith_API
Location[] coord = TestImage(fileName, Model.Hog);
/// Face DetectionWith_DLIB
using (var fd = Dlib.GetFrontalFaceDetector())
{
var img = Dlib.LoadImage <RgbPixel>(fileName);
// find all faces in the image
var faces = fd.Operator(img);
foreach (var face in faces)
{
// draw a rectangle for each face
Dlib.DrawRectangle(img, face, color: new RgbPixel(0, 255, 255), thickness: 4);
}
Dlib.SaveJpeg(img, outputName);
}
// The first thing we are going to do is load all our models. First, since we need to
// find faces in the image we will need a face detector:
using (var detector = Dlib.GetFrontalFaceDetector())
// We will also use a face landmarking model to align faces to a standard pose: (see face_landmark_detection_ex.cpp for an introduction)
using (var sp = ShapePredictor.Deserialize("shape_predictor_68_face_landmarks.dat"))
// And finally we load the DNN responsible for face recognition.
using (var net = DlibDotNet.Dnn.LossMetric.Deserialize("dlib_face_recognition_resnet_model_v1.dat"))
using (var img = Dlib.LoadImageAsMatrix <RgbPixel>(fileName))
using (var win = new ImageWindow(img))
{
var faces = new List <Matrix <RgbPixel> >();
foreach (var face in detector.Operator(img))
{
var shape = sp.Detect(img, face);
var faceChipDetail = Dlib.GetFaceChipDetails(shape, 150, 0.25);
var faceChip = Dlib.ExtractImageChip <RgbPixel>(img, faceChipDetail);
//faces.Add(move(face_chip));
faces.Add(faceChip);
win.AddOverlay(face);
}
if (!faces.Any())
{
Console.WriteLine("No faces found in image!");
return;
}
// This call asks the DNN to convert each face image in faces into a 128D vector.
// In this 128D vector space, images from the same person will be close to each other
// but vectors from different people will be far apart. So we can use these vectors to
// identify if a pair of images are from the same person or from different people.
var faceDescriptors = net.Operator(faces);
// In particular, one simple thing we can do is face clustering. This next bit of code
// creates a graph of connected faces and then uses the Chinese whispers graph clustering
// algorithm to identify how many people there are and which faces belong to whom.
var edges = new List <SamplePair>();
for (uint i = 0; i < faceDescriptors.Count; ++i)
{
for (var j = i; j < faceDescriptors.Count; ++j)
{
// Faces are connected in the graph if they are close enough. Here we check if
// the distance between two face descriptors is less than 0.6, which is the
// decision threshold the network was trained to use. Although you can
// certainly use any other threshold you find useful.
var diff = faceDescriptors[i] - faceDescriptors[j];
if (Dlib.Length(diff) < 0.6)
{
edges.Add(new SamplePair(i, j));
}
}
}
Dlib.ChineseWhispers(edges, 100, out var numClusters, out var labels);
// This will correctly indicate that there are 4 people in the image.
Console.WriteLine($"number of people found in the image: {numClusters}");
// Отобразим результат в ImageList
var winClusters = new List <ImageWindow>();
for (var i = 0; i < numClusters; i++)
{
winClusters.Add(new ImageWindow());
}
var tileImages = new List <Matrix <RgbPixel> >();
for (var clusterId = 0ul; clusterId < numClusters; ++clusterId)
{
var temp = new List <Matrix <RgbPixel> >();
for (var j = 0; j < labels.Length; ++j)
{
if (clusterId == labels[j])
{
temp.Add(faces[j]);
}
}
winClusters[(int)clusterId].Title = $"face cluster {clusterId}";
var tileImage = Dlib.TileImages(temp);
tileImages.Add(tileImage);
winClusters[(int)clusterId].SetImage(tileImage);
}
// Finally, let's print one of the face descriptors to the screen.
using (var trans = Dlib.Trans(faceDescriptors[0]))
{
Console.WriteLine($"face descriptor for one face: {trans}");
// It should also be noted that face recognition accuracy can be improved if jittering
// is used when creating face descriptors. In particular, to get 99.38% on the LFW
// benchmark you need to use the jitter_image() routine to compute the descriptors,
// like so:
var jitterImages = JitterImage(faces[0]).ToArray();
var ret = net.Operator(jitterImages);
using (var m = Dlib.Mat(ret))
using (var faceDescriptor = Dlib.Mean <float>(m))
using (var t = Dlib.Trans(faceDescriptor))
{
Console.WriteLine($"jittered face descriptor for one face: {t}");
// If you use the model without jittering, as we did when clustering the bald guys, it
// gets an accuracy of 99.13% on the LFW benchmark. So jittering makes the whole
// procedure a little more accurate but makes face descriptor calculation slower.
Console.WriteLine("hit enter to terminate");
Console.ReadKey();
foreach (var jitterImage in jitterImages)
{
jitterImage.Dispose();
}
foreach (var tileImage in tileImages)
{
tileImage.Dispose();
}
foreach (var edge in edges)
{
edge.Dispose();
}
foreach (var descriptor in faceDescriptors)
{
descriptor.Dispose();
}
foreach (var face in faces)
{
face.Dispose();
}
}
}
}
System.Console.ReadLine();
}
