Image Segmentation instead of Object Classification #13
Description
Heyo, I have a trained model converted to onxx from pytorch for image segmentation. I can make inference with Python onnx and it is working fine. I can see the segmented classes on the image. However, when I do not know how I can use your code to make the same. I feel like I should use outputTensorValues vector. I tried to convert it to cv::Mat and display it, but it never gives the correct values. I also tried to save the vector to a text file, but all the values are either 0 or 1e-45.