I was wondering if the top view for the Zju-mocap subjects can be generated using this code. I tried using the "rotate_axis" variable, but a proper top view could not be rendered. If I could change the camera center, then possible top views could be generated???
Could you shed some light on what can I possibly look into?
Or
Do I have to create a custom dataset with the top-view camera, then do the smpl for every image, and then render the views??
I am new in this Radiance Field technique, so looking for some guidance.
Thank you for the work @chungyiweng
I was wondering if the top view for the Zju-mocap subjects can be generated using this code. I tried using the "rotate_axis" variable, but a proper top view could not be rendered. If I could change the camera center, then possible top views could be generated???
Could you shed some light on what can I possibly look into?
Or
Do I have to create a custom dataset with the top-view camera, then do the smpl for every image, and then render the views??
I am new in this Radiance Field technique, so looking for some guidance.