You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In the Vanilla 3dgs code, when composing caminfo, fovx and fovy were used instead of the parameter cx and cy in the intrinsics matrix,, I'm a bit confused that it's not implemented on gsplat, as the original implementation did not use cx, cy, How can the camera coordinate system be converted to pixel during rendering? At the same time, when downscale the image, such as factor=2,4, it only adjusts the image size, while cx and cy are not processed. Why is this?
The text was updated successfully, but these errors were encountered:
In the Vanilla 3dgs code, when composing caminfo, fovx and fovy were used instead of the parameter cx and cy in the intrinsics matrix,, I'm a bit confused that it's not implemented on gsplat, as the original implementation did not use cx, cy, How can the camera coordinate system be converted to pixel during rendering? At the same time, when downscale the image, such as factor=2,4, it only adjusts the image size, while cx and cy are not processed. Why is this?
The text was updated successfully, but these errors were encountered: