Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Render specific frames, panoramic views from specific poses, or uniformly sampled from camera path #3251

Open
wants to merge 5 commits into
base: main
Choose a base branch
from

Conversation

paolovic
Copy link
Contributor

@paolovic paolovic commented Jun 23, 2024

Hi,

as described in #2811 I added the following functionality:

  • Enter a desired eye position in world coordinates, and optionally a desired target, and get in return the images of the panoramic view from that pose in parallel to the floor. Anything is adjustable: eye, target, view angle, intermediate steps

    • If you want just a single image, set the start-angle equal to the end-angle and 0 intermediate steps
    • Example usage: ns-render pose-view --output-format images --load-config outputs/room1/splatfacto/2024-06-07_104843/config.yml --eye "0 0 -2" --load-dataparser-transforms outputs/room1/splatfacto/2024-06-07_104843/dataparser_transforms.json
  • Sample uniformly from the camera path and get the views from the corresponding view points

    • Example usage: ns-render pose-view --output-format images --load-config outputs/room1/splatfacto/2024-06-07_104843/config.yml --viewpoints 3

00000
00001
00002
00003
00004
00005
00006
00007
00008
00009
00010

@pwais
Copy link
Contributor

pwais commented Jun 26, 2024

instead of (yet again) wrapping the already-convoluted _render_trajectory_video(), perhaps this could be a script and/or module that outputs a camera_path json file, then the user can just use the existing render.py with that input. and then there's also a camera path fixture that can be (1) unit tested and (2) debugged outside the render loop

@paolovic
Copy link
Contributor Author

instead of (yet again) wrapping the already-convoluted _render_trajectory_video(), perhaps this could be a script and/or module that outputs a camera_path json file, then the user can just use the existing render.py with that input. and then there's also a camera path fixture that can be (1) unit tested and (2) debugged outside the render loop

Hi,
thanks for the feedback, and feedback addressed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants