6DoF Video Point Clouds - Filling in the Gaps 

Bro just wanted you to know youre work is appreciated. One day we will look back at these videos and realize that this is where it all started. Youre an innovator for true volumetric immersive video using one camera. A true pioneer in this field. Keep up the good work sir!!

Hello, So discovering 6dof video exists kinda blow my mind, I've been trying to build 3d point clouds from single camera video like forever, SFM works but only for specific type of takes, and for other dolly like takes you need a 3d camera tracker like PFtrack which is super expensive and does not work perfectly. In regards to the topic of the video, I think a good way to fill in the gaps would be, since you've a point cloud, to look in previous/later frames, for points that fall in that shaded area, lets say, the ground, the ground that in an specific moment is ocluded it may be visible in other frames, so, feel in those frames with points from other frames.

This is pretty awesome. It feels like there must be a method of generating a mesh from the sum of all points in the video where the motion of the points is below a certain threshold. Or perhaps it's something like you take the points with the furthest depth value and use those to generate and texture a mesh plate.

This is a great starting point to dive into this stuff, cheers and thanks! How are you generating your point clouds? I've been trying out various software without any usable results for Unity.