/// <summary>
/// Parses net output (detect) to predictions.
/// </summary>
private List <YoloPrediction> ParseDetect(DenseTensor <float> output, Image image)
{
var result = new ConcurrentBag <YoloPrediction>();
var(w, h) = (image.Width, image.Height); // image w and h
var(xGain, yGain) = (_model.Width / (float)w, _model.Height / (float)h); // x, y gains
var gain = Math.Min(xGain, yGain); // gain = resized / original
var(xPad, yPad) = ((_model.Width - w * gain) / 2, (_model.Height - h * gain) / 2); // left, right pads
Parallel.For(0, (int)output.Length / _model.Dimensions, (i) =>
{
if (output[0, i, 4] <= _model.Confidence)
{
return; // skip low obj_conf results
}
Parallel.For(5, _model.Dimensions, (j) =>
{
output[0, i, j] = output[0, i, j] * output[0, i, 4]; // mul_conf = obj_conf * cls_conf
});
Parallel.For(5, _model.Dimensions, (k) =>
{
if (output[0, i, k] <= _model.MulConfidence)
{
return; // skip low mul_conf results
}
float xMin = ((output[0, i, 0] - output[0, i, 2] / 2) - xPad) / gain; // unpad bbox tlx to original
float yMin = ((output[0, i, 1] - output[0, i, 3] / 2) - yPad) / gain; // unpad bbox tly to original
float xMax = ((output[0, i, 0] + output[0, i, 2] / 2) - xPad) / gain; // unpad bbox brx to original
float yMax = ((output[0, i, 1] + output[0, i, 3] / 2) - yPad) / gain; // unpad bbox bry to original
xMin = Clamp(xMin, 0, w - 0); // clip bbox tlx to boundaries
yMin = Clamp(yMin, 0, h - 0); // clip bbox tly to boundaries
xMax = Clamp(xMax, 0, w - 1); // clip bbox brx to boundaries
yMax = Clamp(yMax, 0, h - 1); // clip bbox bry to boundaries
YoloLabel label = _model.Labels[k - 5];
var prediction = new YoloPrediction(label, output[0, i, k])
{
Rectangle = new RectangleF(xMin, yMin, xMax - xMin, yMax - yMin)
};
result.Add(prediction);
});
});
return(result.ToList());
}
/// <summary>
/// Parses net output (detect) to predictions.
/// </summary>
private List <YoloPrediction> ParseDetect(DenseTensor <float> output, Image image)
{
var result = new List <YoloPrediction>();
var(xGain, yGain) = (_model.Width / (float)image.Width, _model.Height / (float)image.Height);
for (int i = 0; i < output.Length / _model.Dimensions; i++) // iterate tensor
{
if (output[0, i, 4] <= _model.Confidence)
{
continue;
}
for (int j = 5; j < _model.Dimensions; j++) // compute mul conf
{
output[0, i, j] = output[0, i, j] * output[0, i, 4]; // conf = obj_conf * cls_conf
}
for (int k = 5; k < _model.Dimensions; k++)
{
if (output[0, i, k] <= _model.MulConfidence)
{
continue;
}
var xMin = (output[0, i, 0] - output[0, i, 2] / 2) / xGain; // top left x
var yMin = (output[0, i, 1] - output[0, i, 3] / 2) / yGain; // top left y
var xMax = (output[0, i, 0] + output[0, i, 2] / 2) / xGain; // bottom right x
var yMax = (output[0, i, 1] + output[0, i, 3] / 2) / yGain; // bottom right y
YoloLabel label = _model.Labels[k - 5];
var prediction = new YoloPrediction(label, output[0, i, k])
{
Rectangle = new RectangleF(xMin, yMin, xMax - xMin, yMax - yMin)
};
result.Add(prediction);
}
}
return(result);
}
/// <summary>
/// Parses net outputs (sigmoid) to predictions.
/// </summary>
private List <YoloPrediction> ParseSigmoid(DenseTensor <float>[] output, Image image)
{
var result = new ConcurrentBag <YoloPrediction>();
var(w, h) = (image.Width, image.Height); // image w and h
var(xGain, yGain) = (_model.Width / (float)w, _model.Height / (float)h); // x, y gains
var gain = Math.Min(xGain, yGain); // gain = resized / original
var(xPad, yPad) = ((_model.Width - w * gain) / 2, (_model.Height - h * gain) / 2); // left, right pads
Parallel.For(0, output.Length, (i) => // iterate model outputs
{
int shapes = _model.Shapes[i]; // shapes per output
Parallel.For(0, _model.Anchors[0].Length, (a) => // iterate anchors
{
Parallel.For(0, shapes, (y) => // iterate shapes (rows)
{
Parallel.For(0, shapes, (x) => // iterate shapes (columns)
{
int offset = (shapes * shapes * a + shapes * y + x) * _model.Dimensions;
float[] buffer = output[i].Skip(offset).Take(_model.Dimensions).Select(Sigmoid).ToArray();
if (buffer[4] <= _model.Confidence)
{
return; // skip low obj_conf results
}
List <float> scores = buffer.Skip(5).Select(b => b * buffer[4]).ToList(); // mul_conf = obj_conf * cls_conf
float mulConfidence = scores.Max(); // max confidence score
if (mulConfidence <= _model.MulConfidence)
{
return; // skip low mul_conf results
}
float rawX = (buffer[0] * 2 - 0.5f + x) * _model.Strides[i]; // predicted bbox x (center)
float rawY = (buffer[1] * 2 - 0.5f + y) * _model.Strides[i]; // predicted bbox y (center)
float rawW = (float)Math.Pow(buffer[2] * 2, 2) * _model.Anchors[i][a][0]; // predicted bbox w
float rawH = (float)Math.Pow(buffer[3] * 2, 2) * _model.Anchors[i][a][1]; // predicted bbox h
float[] xyxy = Xywh2xyxy(new float[] { rawX, rawY, rawW, rawH });
float xMin = Clamp((xyxy[0] - xPad) / gain, 0, w - 0); // unpad, clip tlx
float yMin = Clamp((xyxy[1] - yPad) / gain, 0, h - 0); // unpad, clip tly
float xMax = Clamp((xyxy[2] - xPad) / gain, 0, w - 1); // unpad, clip brx
float yMax = Clamp((xyxy[3] - yPad) / gain, 0, h - 1); // unpad, clip bry
YoloLabel label = _model.Labels[scores.IndexOf(mulConfidence)];
var prediction = new YoloPrediction(label, mulConfidence)
{
Rectangle = new RectangleF(xMin, yMin, xMax - xMin, yMax - yMin)
};
result.Add(prediction);
});
});
});
});
return(result.ToList());
}
/// <summary>
/// Parses net output to predictions.
/// </summary>
private List <YoloPrediction> ParseOutput(DenseTensor <float>[] output, Image image)
{
var result = new List <YoloPrediction>();
var(xGain, yGain) = (_model.Width / (float)image.Width, _model.Height / (float)image.Height);
for (int i = 0; i < output.Length; i++) // iterate outputs
{
int shapes = _shapes[i]; // shapes per output
for (int a = 0; a < _anchors.Length; a++) // iterate anchors
{
for (int y = 0; y < shapes; y++) // iterate rows
{
for (int x = 0; x < shapes; x++) // iterate columns
{
int offset = (shapes * shapes * a + shapes * y + x) * _model.Dimensions;
float[] buffer = output[i].Skip(offset).Take(_model.Dimensions).Select(Sigmoid).ToArray();
var objConfidence = buffer[4]; // extract object confidence
if (objConfidence < _model.Confidence) // check predicted object confidence
{
continue;
}
List <float> scores = buffer.Skip(5).Select(x => x * objConfidence).ToList();
float mulConfidence = scores.Max(); // find the best label
if (mulConfidence <= _model.MulConfidence) // check class obj_conf * cls_conf confidence
{
continue;
}
var rawX = (buffer[0] * 2 - 0.5f + x) * _strides[i]; // predicted bbox x (center)
var rawY = (buffer[1] * 2 - 0.5f + y) * _strides[i]; // predicted bbox y (center)
var rawW = MathF.Pow(buffer[2] * 2, 2) * _anchors[i][a][0]; // predicted bbox width
var rawH = MathF.Pow(buffer[3] * 2, 2) * _anchors[i][a][1]; // predicted bbox height
float[] xyxy = Xywh2xyxy(new float[] { rawX, rawY, rawW, rawH });
var xMin = xyxy[0] / xGain; // final bbox tlx scaled with ratio (to original size)
var yMin = xyxy[1] / yGain; // final bbox tly scaled with ratio (to original size)
var xMax = xyxy[2] / xGain; // final bbox brx scaled with ratio (to original size)
var yMax = xyxy[3] / yGain; // final bbox bry scaled with ratio (to original size)
YoloLabel label = _model.Labels[scores.IndexOf(mulConfidence)];
var prediction = new YoloPrediction(label, mulConfidence)
{
Rectangle = new RectangleF(xMin, yMin, xMax - xMin, yMax - yMin)
};
result.Add(prediction);
}
}
}
}
return(result);
}