search

autonomousvision / monosdf

[NeurIPS'22] MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction
MIT License
573 stars 53 forks source link

question about how to make the video clip #47

Closed fangchuan closed 2 years ago

fangchuan commented 2 years ago

Hi, thanks for open sourcing this amazing work. I have test the code and the result is fine. I have a non-algorithmic question for you. You showed the awesome mesh reconstruction results on the project page. I would like to know what tools you used to render that video. The input was mesh and the output was a video clip. @niujinshuchong

niujinshuchong commented 2 years ago

Hi, we used open3d for rendering.

fangchuan commented 2 years ago

Hi, we used open3d for rendering.

Great, I also use open3d to render images from the resulting mesh, but I fail to copy with the appropriate illumination and shadow, so the output video is not good enough, Could you share your rendering code snippet? I want to check if there is anything I get wrong with the open3d.visualization settings.

niujinshuchong commented 2 years ago

Hi, the script for rendering is uploaded here: https://github.com/autonomousvision/monosdf/blob/main/render/render_tntvideos_open3d.py

fangchuan commented 2 years ago

Hi, the script for rendering is uploaded here: https://github.com/autonomousvision/monosdf/blob/main/render/render_tntvideos_open3d.py

Thanks, I really appreciate your help!