fatPeter / mini-splatting

103 stars 10 forks source link

About Visualization #4

Open wikiwa opened 3 months ago

wikiwa commented 3 months ago

330293308-651a6489-4ff6-42e9-9863-8371bcca0748

Excellent work, I would like to know how you create visualizations, specifically how you achieve the visualizations in your charts.

fatPeter commented 3 months ago

Thank you for your interest. The script (teaser.py) to draw our teaser has been uploaded. Use it as render.py:

# teaser, add other params as render.py
python teaser.py -s <dataset path> -m <model path>
zhangseammm commented 1 month ago

Great work! I run your teaser.py and do you know how to solve this problem? I run render.py with same parameters and inputs, it works.

Error: no "view" rule for type "image/png" passed its test case (for more information, add "--debug=1" on the command line)

fatPeter commented 1 month ago

This script works well on my pc. It might be a warning (rather than error) from PIL image.show when running on a server. Simply add image_proj.save(r'./output.jpg') to save the rendered image as:

    # tune point size for better visualization 0.3, 0.3, 1.2
    image_proj = draw_points_on_image(points, np.zeros(colors.shape)+[0,0,255], rendering.permute(1,2,0).detach().cpu().numpy(), size=0.3)
    image_proj.save(r'./output.jpg')

    return 

Rendering result on our server: image

zhangseammm commented 1 month ago

Thank you soooo much!