Custom model and problem.

[magicsoft75]

Hi.
Thank you for the solution - it looks good :-)
Unfortunately, I have a problem.
I trained a model (4 classes), converted to onnx - works with detect.py but doesn't work with Yolov5Net.
Model input/output: images/output0.
I changed Dimmensions to 9 - what else do I have to change?
Thank you

[magicsoft75]

Ok, I have it ;-)
In file YoloCocoP5Model.cs we have to change:
public override int Dimensions { get; set; } = 9; //Number of class + 5
public override List Labels { get; set; } = new List()
{ to own labels
and
public override string[] Outputs { get; set; } = new[] { "output0" };