data
biyuefeng opened this issue · 3 comments
biyuefeng commented
I only have images and masks, how should I train them? Thank you.
Cjl-MedSeg commented
We provide training examples in data_demo, and you can refer to their data format to build your own training data to start training.
biyuefeng commented
I annotated the data using labelme, which includes images and JSON files. I generated a mask using JSON. However, there may be multiple mask targets on one of my images. I have seen your data, and the mask image below each image contains a target. How can I organize these mask images in your format? Thank you!
…---Original---
From: ***@***.***>
Date: Wed, Jan 17, 2024 23:55 PM
To: ***@***.***>;
Cc: ***@***.******@***.***>;
Subject: Re: [OpenGVLab/SAM-Med2D] data (Issue #47)
We provide training examples in data_demo, and you can refer to their data format to build your own training data to start training.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
biyuefeng commented
Hello, I encountered the following error while training my dataset:
I have added an attachment to the JSON file after converting the dataset. Hope to be resolved, thank you!
人生海海海布里
***@***.***
…------------------ 原始邮件 ------------------
发件人: "OpenGVLab/SAM-Med2D" ***@***.***>;
发送时间: 2024年1月17日(星期三) 晚上11:55
***@***.***>;
***@***.******@***.***>;
主题: Re: [OpenGVLab/SAM-Med2D] data (Issue #47)
We provide training examples in data_demo, and you can refer to their data format to build your own training data to start training.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>