mdciri / yolov7-bone-fracture-detection Goto Github PK

I was testing the model that you produced and it is really great. Today I found some images in kaggle and tried to test the model again but I was surprised that it didn't produce a good result. could you test the image and check, I might be doing something wrong.

Hi Dear,

please if it is possible to describe the steps of training the model in the local machine

It seems there is no train.py file

How convert pt model to onnx model

theres no requirements.txt , should i just use the docker requirements?

First of all great work. Secondly I tested inference on random images but sometimes it fails to detect clear fractures, is it possible that algorithm prefers certain dimensions, size or resolution of images? Or even certain types of bones found in body?