I had to use Yolov5 in a project of mine. This dataset gave better results than the Yolov5 in my case. So, took the x6 and trained in on this dataset. Attached is a drive link to the README.md since GitHub doesn't allow files larger than 100MB and LFS couldn't be used on a forked project.
I had to use Yolov5 in a project of mine. This dataset gave better results than the Yolov5 in my case. So, took the x6 and trained in on this dataset. Attached is a drive link to the README.md since GitHub doesn't allow files larger than 100MB and LFS couldn't be used on a forked project.