How to implement sliding window training？ #4268
-
|
Thanks for releasing such a great job! I found that sliding window inference has been implemented ,so how to implement sliding window training? |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments
-
|
Generally speaking you don't want to use sliding window inference for training, it's non-differentiable. Typically we use random crops for training and then use sliding window inference for validation/testing/deployment. |
Beta Was this translation helpful? Give feedback.
-
|
@lanslotttTT can you share more details about sliding-window training? is it similar to the sliding-window scheme in swin transformer? |
Beta Was this translation helpful? Give feedback.
-
|
I mean it's a normal method，like sliding window inference.
…---- 回复的原邮件 ----
| 发件人 | ***@***.***> |
| 日期 | 2022年05月14日 02:39 |
| 收件人 | ***@***.***> |
| 抄送至 | ***@***.******@***.***> |
| 主题 | Re: [Project-MONAI/MONAI] How to implement sliding window training？ (Discussion #4268) |
@lanslotttTT can you share more details about sliding-window training? is it similar to the sliding-window scheme in swin transformer?
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
Generally speaking you don't want to use sliding window inference for training, it's non-differentiable. Typically we use random crops for training and then use sliding window inference for validation/testing/deployment.