How to use yolov3 onnx model for image object detection with microsoft.ml (ML.NET 1.4)

Question

0.00/5 (No votes)

See more:

, +

Hello,
i have a question about yolov3
when i use ML.NET 1.4 for image object detection

What I have tried:

i study ML.NET sample:https://github.com/dotnet/machinelearning-samples/tree/master/samples/csharp/getting-started/DeepLearning_ObjectDetection_Onnx

and modify pipline for yolov3 input and ouput

C#

var pipeline = mlContext.Transforms.LoadImages("input_1", "", nameof(ImageNetData.ImagePath))
                .Append(mlContext.Transforms.ResizeImages("input_1", ImageNetSettings.imageWidth, ImageNetSettings.imageHeight, "input_1"))
                .Append(mlContext.Transforms.ExtractPixels("input_1"))
                .Append(mlContext.Transforms.ApplyOnnxModel(modelFile: modelLocation, outputColumnNames: new[]
                {
                    "yolonms_layer_1/ExpandDims_1:0",
                    "yolonms_layer_1/ExpandDims_3:0",
                    "yolonms_layer_1/concat_2:0"
                }, inputColumnNames: new[] {"input_1"}));

when i predict data, i dont know how to modify three output ids,
because this method only accept one input arg

C#

var scoredData = model.Transform(testData);

            IEnumerable<float[]> probabilities =
                scoredData.GetColumn<float[]>("yolonms_layer_1/concat_2:0").ToList();

yolov3 onnx model download from https://github.com/onnx/models
and use Netron to get output layers

please give me some guidance, thank you all very much!

Posted 30-Nov-19 15:15pm

mx117eva007

Updated 24-Oct-20 1:09am

v6

Add a Solution

3 solutions

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Member 14685145 · Answer 1 · 2019-12-10T10:46:00

Solution 1

Hello, I faced the same issue as you...
Did you find a solution for it?

Posted 10-Dec-19 10:46am

Member 14685145

Comments

mx117eva007 15-Dec-19 2:21am

i didn't find the solution ...

REZIDENT10 · Answer 2 · 2020-06-05T08:44:00

.Append(mlContext.Transforms.ApplyOnnxModel(modelFile: modelLocation, outputColumnNames: new[] { "yolonms_layer_1/ExpandDims_1:0",
"yolonms_layer_1/ExpandDims_3:0",
"yolonms_layer_1/concat_2:0" }, inputColumnNames: new[] { "input_1" }))
.Append(mlContext.Transforms.Concatenate("outputs", new[] { "yolonms_layer_1/ExpandDims_1:0",
"yolonms_layer_1/ExpandDims_3:0",
"yolonms_layer_1/concat_2:0" }));

IEnumerable<float[]> probabilities =
scoredData.GetColumn<float[]>("outputs").ToList();

Member 14974057 · Answer 3 · 2020-10-24T01:09:00

Solution 3

see How to impelement post-proccesing for yolo v3 or v4 onnx models in ML.Net - Stack Overflow[^]

Posted 24-Oct-20 1:09am

Member 14974057