Hiya,
So I have recently made an ONNX web application for object detection and now moving to segmentation.
I have followed the crack-seg on the ultralytics website and have converted it to an ONNX model (with an input shape of 320*320).
I am now just stuck trying to interpret the output data.
Below is the picture of the output.
I understand the first 10500 numbers on the output0 (the bounding box x, y, width and height then the confidence of the one class)
I now have no clue how to understand what output1 is, i imagine it is 32 lots of 80 x and y positions but just wonder if any one else knows what it could be.
Thanks!