Skip to content

editor.md

Latest commit

 

History

History
76 lines (50 loc) · 2.75 KB

editor.md

File metadata and controls

76 lines (50 loc) · 2.75 KB

Viewer and Editor

Our viewer and editor is based on viser

Selected videos rendered by our editor, Please note: These videos are encoded using HEVC with 10-bit HDR colors and are best viewed on a compatible display with HDR support, e.g. recent Apple devices.

windowlegovary.mp4
gardenlights.mp4

Editor

After training, you could use the following command to view and edit the scene!

python basicgs/view.py output/le3d/windowlegovary --edit

Like most of the editors, you could use key frames and interpolations to create your own video storyboard!

edit_demo.mp4

Just set the KeyFrames! All the interpolation will be done automatically! (BTW, you could set the acceleration of the camera motion in the Interpolations!)

Viewer

View During Training

If you turn the viewer option in the yaml option on,

viewer:
  type: Le3dViewer
  update_freq: 1
  port: 8097

and you could see this in the log.

INFO:websockets.server:server listening on 0.0.0.0:8097
╭──────────────── viser ────────────────╮
│             ╷                         │
│   HTTP      │ http://localhost:8097   │
│   Websocket │ ws://localhost:8097     │
│             ╵                         │
╰───────────────────────────────────────╯
2025-01-08 06:29:35,314 INFO: Viewer initialized, port: 8097

You could open the http://localhost:8097 in your browser to view the scene.
Here is the screenshot of the viewer, where contains the post-processing, the camera pose, and the training losses.

image

And you could also render the depth, final opacity, and the histogram of the scene.

2025-01-08.14.40.17.mov

View After Training

First got your trained model ready, and then you could use the following command to view the scene!

python basicgs/view.py output/le3d/windowlegovary

You could change the exposure, hdr tone mapping, camera pose, defocus blur, and even curve enhancement on that!

view_demo.mp4