Hi, the option to download as “Yolov8 Image Segmentation” option is not turning up in the download function for my above project that is linked. It only shows these formats in the download tab (Image below) and when I scroll down, other options are not “Yolov8 Image Segmentation” either - the closest is “Yolov8”:
Additionally, I have tried to fix this issue by downloading a zip file with the format of “COCO Segmentation” and uploaded this to a 2nd newly created project that was set from the start as Image Segmentation project (in case the reason the first project failed was because I set it as Object Detection at first then changed it to Image Segmentation), but the download tab still showed the same few options.
Besides the download function that I have been talking about above, I am also unable to locate the export function even after searching everywhere for it. Has this been removed?
How can I fix this, and any help will be appreciated!
Hi @Lau_En_Xin!
With regards to your first question, we are currently investigating and I will update this post once we have a resolution.
With regards to your second question, assuming by export you mean fork, I have created a guide below to walk you through the process. This allows you to fork the dataset into your own Roboflow project.
1) Navigate to the Overview tab under General. Click on Fork Project (Circled in red below)
2) You will arrive to the screen below. Choose your Roboflow Project that you would like to Fork the Dataset into and select Fork Dataset (Encircled in red below).
Thank you very much for responding to my question. I was wondering if there are any updates on this issue, as I exported this dataset as “COCO Image Segmentation”, trained the model in a virtual environment on my local computer on python 3.8, and deployed in on Jetson Nano (4GB) on an image that I used to train the very model. But the output image had bounding boxes clustered on the top left corner, the boxes were all tiny and the labels were completely wrong (gave “carb” instead of the correct class “meat”).
Hence it would be great if I was able to solve this problem by exporting it as it was meant to be exported - as “Yolov8n Image Segmentation”.
