Open SlamCabbage opened 2 months ago
torch.exp is used in build_scaling_rotation, so is the "scales" saved at the end the actual size of the Gaussian in this scene?
Hi, when I refers to the scales, I mean the post-activation scales. Yes, the scales are indeed measured in the actual scene; that is, the world space. You need to append the scale-post-activation for recovering a full covariance.
大佬您好,我用中文了,这是waymo数据,我可视化它的协方差,这个是按照scale的0.05倍可视化的。如果按照1.0倍进行可视化,那么整个场景都是一坨纠缠在一起的,完全看不出来。
Waymo should not be a problem. https://streetunveiler.github.io/ I use 2DGS in my project above. And I at least successfully reconstruct the scene.
我对scales做了一个exp操作之后,可视化的结果如上图。
用2d gs可视化工具能看到这个场景的地面点虽然不多,但是在图像上是覆盖全了,为什么我对高斯进行可视化,地面并没有能够把空出补全的大高斯平面呢?
Waymo should not be a problem. https://streetunveiler.github.io/ I use 2DGS in my project above. And I at least successfully reconstruct the scene.
Already following, looking forward to open source
padding zeros to the scales and pass them into
build_scaling_rotation