Skip to content

Commit

Permalink
start implementing yolov5l
Browse files Browse the repository at this point in the history
  • Loading branch information
@BobLd
BobLd committed Dec 8, 2020
1 parent 5de7e16 commit c9e1f42
Show file tree
Hide file tree
Showing 2 changed files with 57 additions and 40 deletions.
75 changes: 46 additions & 29 deletions YOLOv4MLNet/DataStructures/YoloV4Prediction.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -11,41 +11,41 @@ public class YoloV4Prediction
// https://github.com/hunglc007/tensorflow-yolov4-tflite/blob/master/data/anchors/yolov4_anchors.txt
static readonly float[][][] ANCHORS = new float[][][]
{
new float[][] { new float[] { 12, 16 }, new float[] { 19, 36 }, new float[] { 40, 28 } },
new float[][] { new float[] { 36, 75 }, new float[] { 76, 55 }, new float[] { 72, 146 } },
new float[][] { new float[] { 142, 110 }, new float[] { 192, 243 }, new float[] { 459, 401 } }
new float[][] { new float[] { 10, 13 }, new float[] { 16, 30 }, new float[] { 33, 23 } },
new float[][] { new float[] { 30, 61 }, new float[] { 62, 45 }, new float[] { 59, 119 } },
new float[][] { new float[] { 116, 90 }, new float[] { 156, 198 }, new float[] { 373, 326 } }
};

// https://github.com/hunglc007/tensorflow-yolov4-tflite/blob/9f16748aa3f45ff240608da4bd9b1216a29127f5/core/config.py#L18
static readonly float[] STRIDES = new float[] { 8, 16, 32 };
static readonly float[] STRIDES = new float[] { 8, 16, 32 }; // 640/80, 640/40, 640/20

// https://github.com/hunglc007/tensorflow-yolov4-tflite/blob/9f16748aa3f45ff240608da4bd9b1216a29127f5/core/config.py#L20
static readonly float[] XYSCALE = new float[] { 1.2f, 1.1f, 1.05f };
static readonly float[] XYSCALE = new float[] { 1f, 1f, 1f };

static readonly int[] shapes = new int[] { 52, 26, 13 };
static readonly int[] shapes = new int[] { 80, 40, 20 };

const int anchorsCount = 3;

/// <summary>
/// Identity
/// </summary>
[VectorType(1, 52, 52, 3, 85)]
[ColumnName("Identity:0")]
public float[] Identity { get; set; }
[VectorType(1, 3, 80, 80, 85)]
[ColumnName("output")]
public float[] Output { get; set; }

/// <summary>
/// Identity1
/// </summary>
[VectorType(1, 26, 26, 3, 85)]
[ColumnName("Identity_1:0")]
public float[] Identity1 { get; set; }
[VectorType(1, 3, 40, 40, 85)]
[ColumnName("1313")]
public float[] Output1313 { get; set; }

/// <summary>
/// Identity2
/// </summary>
[VectorType(1, 13, 13, 3, 85)]
[ColumnName("Identity_2:0")]
public float[] Identity2 { get; set; }
[VectorType(1, 3, 20, 20, 85)]
[ColumnName("1333")]
public float[] Output1333 { get; set; }

[ColumnName("width")]
public float ImageWidth { get; set; }
Expand All @@ -57,23 +57,24 @@ public IReadOnlyList<YoloV4Result> GetResults(string[] categories, float scoreTh
{
List<float[]> postProcesssedResults = new List<float[]>();
int classesCount = categories.Length;
var results = new[] { Identity, Identity1, Identity2 };
var results = new[] { Output, Output1313, Output1333 };

for (int i = 0; i < results.Length; i++)
{
var pred = results[i];
var outputSize = shapes[i];

for (int boxY = 0; boxY < outputSize; boxY++)
{
for (int boxX = 0; boxX < outputSize; boxX++)
{
for (int a = 0; a < anchorsCount; a++)
{
var offset = (boxY * outputSize * (classesCount + 5) * anchorsCount) + (boxX * (classesCount + 5) * anchorsCount) + a * (classesCount + 5);
var predBbox = pred.Skip(offset).Take(classesCount + 5).ToArray();
var predBbox = pred.Skip(offset).Take(classesCount + 5).Select(x => Sigmoid(x)).ToArray(); // y = x[i].sigmoid()

// ported from https://github.com/onnx/models/tree/master/vision/object_detection_segmentation/yolov4#postprocessing-steps
// more info at
// https://github.com/ultralytics/yolov5/issues/343#issuecomment-658021043
// https://github.com/ultralytics/yolov5/blob/a1c8406af3eac3e20d4dd5d327fd6cbd4fbb9752/models/yolo.py#L29-L36

// postprocess_bbbox()
var predXywh = predBbox.Take(4).ToArray();
Expand All @@ -85,23 +86,24 @@ public IReadOnlyList<YoloV4Result> GetResults(string[] categories, float scoreTh
var rawDw = predXywh[2];
var rawDh = predXywh[3];

float predX = ((Sigmoid(rawDx) * XYSCALE[i]) - 0.5f * (XYSCALE[i] - 1) + boxX) * STRIDES[i];
float predY = ((Sigmoid(rawDy) * XYSCALE[i]) - 0.5f * (XYSCALE[i] - 1) + boxY) * STRIDES[i];
float predW = (float)Math.Exp(rawDw) * ANCHORS[i][a][0];
float predH = (float)Math.Exp(rawDh) * ANCHORS[i][a][1];
float predX = ((rawDx * 2f) - 0.5f + boxX) * STRIDES[i];
float predY = ((rawDy * 2f) - 0.5f + boxY) * STRIDES[i];
float predW = (float)Math.Pow(rawDw * 2, 2) * ANCHORS[i][a][0];
float predH = (float)Math.Pow(rawDh * 2, 2) * ANCHORS[i][a][1];

// postprocess_boxes
// (1) (x, y, w, h) --> (xmin, ymin, xmax, ymax)
float predX1 = predX - predW * 0.5f;
float predY1 = predY - predH * 0.5f;
float predX2 = predX + predW * 0.5f;
float predY2 = predY + predH * 0.5f;
var box = Xywh2xyxy(new float[] { predX, predY, predW, predH });
float predX1 = box[0]; //predX - predW * 0.5f;
float predY1 = box[1]; //predY - predH * 0.5f;
float predX2 = box[2]; //predX + predW * 0.5f;
float predY2 = box[3]; //predY + predH * 0.5f;

// (2) (xmin, ymin, xmax, ymax) -> (xmin_org, ymin_org, xmax_org, ymax_org)
float org_h = ImageHeight;
float org_w = ImageWidth;

float inputSize = 416f;
float inputSize = 640f;
float resizeRatio = Math.Min(inputSize / org_w, inputSize / org_h);
float dw = (inputSize - resizeRatio * org_w) / 2f;
float dh = (inputSize - resizeRatio * org_h) / 2f;
Expand Down Expand Up @@ -160,7 +162,7 @@ public IReadOnlyList<YoloV4Result> GetResults(string[] categories, float scoreTh
if (float.IsNaN(iou[i])) continue;
if (iou[i] > iouThres)
{
postProcesssedResults[i] = null;
//postProcesssedResults[i] = null; // deactivated for debugging
}
}
f++;
Expand Down Expand Up @@ -200,5 +202,20 @@ static float box_area(float[] box)

return inter / (area1 + area2 - inter);
}

/// <summary>
/// Convert bounding box format from [x, y, w, h] to [x1, y1, x2, y2]
/// <para>Box (center x, center y, width, height) to (x1, y1, x2, y2)</para>
/// </summary>
public static float[] Xywh2xyxy(float[] bbox)
{
//https://github.com/BobLd/YOLOv3MLNet/blob/48d691175249c2dbf1fdfb9790c9aec56f9a028f/YOLOv3MLNet/DataStructures/YoloV3Prediction.cs#L159-L167
var bboxAdj = new float[4];
bboxAdj[0] = bbox[0] - bbox[2] / 2f;
bboxAdj[1] = bbox[1] - bbox[3] / 2f;
bboxAdj[2] = bbox[0] + bbox[2] / 2f;
bboxAdj[3] = bbox[1] + bbox[3] / 2f;
return bboxAdj;
}
}
}
22 changes: 11 additions & 11 deletions YOLOv4MLNet/Program.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ class Program
{
// model is available here:
// https://github.com/onnx/models/tree/master/vision/object_detection_segmentation/yolov4
const string modelPath = @"D:\MachineLearning\Models\yolo_models\yolov4.onnx";
const string modelPath = @"D:\MachineLearning\Models\yolo_models\yolov5l.onnx";

const string imageFolder = @"Assets\Images";

Expand All @@ -29,25 +29,25 @@ static void Main()
// https://github.com/onnx/models/tree/master/vision/object_detection_segmentation/yolov4

// Define scoring pipeline
var pipeline = mlContext.Transforms.ResizeImages(inputColumnName: "bitmap", outputColumnName: "input_1:0", imageWidth: 416, imageHeight: 416, resizing: ResizingKind.IsoPad)
.Append(mlContext.Transforms.ExtractPixels(outputColumnName: "input_1:0", scaleImage: 1f / 255f, interleavePixelColors: true))
var pipeline = mlContext.Transforms.ResizeImages(inputColumnName: "bitmap", outputColumnName: "images", imageWidth: 640, imageHeight: 640, resizing: ResizingKind.IsoPad)
.Append(mlContext.Transforms.ExtractPixels(outputColumnName: "images", scaleImage: 1f / 255f, interleavePixelColors: false))
.Append(mlContext.Transforms.ApplyOnnxModel(
shapeDictionary: new Dictionary<string, int[]>()
{
{ "input_1:0", new[] { 1, 416, 416, 3 } },
{ "Identity:0", new[] { 1, 52, 52, 3, 85 } },
{ "Identity_1:0", new[] { 1, 26, 26, 3, 85 } },
{ "Identity_2:0", new[] { 1, 13, 13, 3, 85 } },
{ "images", new[] { 1, 3, 640, 640 } },
{ "output", new[] { 1, 3, 80, 80, 85 } },
{ "1313", new[] { 1, 3, 40, 40, 85 } },
{ "1333", new[] { 1, 3, 20, 20, 85 } },
},
inputColumnNames: new[]
{
"input_1:0"
"images"
},
outputColumnNames: new[]
{
"Identity:0",
"Identity_1:0",
"Identity_2:0"
"output",
"1313",
"1333"
},
modelFile: modelPath));

Expand Down

0 comments on commit c9e1f42

Please sign in to comment.