Skip to content
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

关于sparsebev_sampling中的sample_points_cam以及5-D tensor采样问题 #74

Open
kingofstu opened this issue Jun 5, 2024 · 1 comment

Comments

@kingofstu
Copy link

kingofstu commented Jun 5, 2024

作者您好,我有两个疑问希望可以被解答
1、我看到sparsebev_sampling的代码中并未使用valid_mask,也就是说sample_points_cam中是存在投影不到相机平面的3d reference_points的,这样采样的时候不会有影响吗?padding会填充0对吗?
2、使用i_view[..., None].float() / (N - 1)拼接到sample_points_cam是因为each muvl_feats:[BTG, N, C, H, W]是5-D tensor,因此要cat相机索引到采样点中然后进行采样吗?那如果将each muvl_feats变为[BTG*N, C, H, W]是不是可以直接用4-D tensor了呢?

源码:
i_view = torch.argmax(valid_mask, dim=-1)[..., None] # [B, T, Q, GP, 1] 对于每个3d点,只取其在一个相机的投影
sample_points_cam = sample_points_cam[i_batch, i_time, i_query, i_point, i_view, :] # [B, Q, GP, 1, 2] 这里存在投射出图像平面外的3d点的投射坐标
valid_mask = valid_mask[i_batch, i_time, i_query, i_point, i_view] # [B, Q, GP, 1]
sample_points_cam = torch.cat([sample_points_cam, i_view[..., None].float() / (N - 1)], dim=-1)

@rubbish001
Copy link

你调试一下就知道了,无效的点的特征都是0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants