-
Notifications
You must be signed in to change notification settings - Fork 1.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Feature] Add HRNet_W48_Contrast Model #1484
[Feature] Add HRNet_W48_Contrast Model #1484
Conversation
configs/hrnet_w48_contrast/HRNet_W48_contrast_cityscapes_1024x512_60k.yml
Outdated
Show resolved
Hide resolved
precommit 修改格式:#1490 (comment) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
所有要求修改已经完成,我会重新训练达到要求精度之后进行合入。Good job~
paddleseg/models/hrnet_contrast.py
Outdated
Args: | ||
dim_in (int): The dimensions of input features. | ||
proj_dim (int|optional): The output dimensions of projection head. Default: 256. | ||
proj (str|optional): The type of projection head, only support 'linear' and 'convmlp'. Default: 'convmlp' |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Default: 'convmlp'.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@shiyutang 已review完,验证导出功能后可以合入
添加Exploring Cross-Image Pixel Contrast for Semantic Segmentation论文的模型到paddleseg中,并添加了对应的config文件。
1、添加model,HRNetW48Contrast
2、添加PixelContrastCrossEntropyLoss损失函数