Lay3rss / SuGaR

Forked repo of "Official PyTorch implementation of SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering"
https://anttwo.github.io/sugar/
Other
0 stars 0 forks source link

Bureau - Hotel de la Marine - H100 #4

Open rebecca-lay3rs opened 4 months ago

rebecca-lay3rs commented 4 months ago

1. Training using train.py (calls the scrips automatically)

1.1 Align cameras using COLMAP implementation of GS

python gaussian_splatting/convert.py -s data/Bureau/

1.2 Train GS model

python gaussian_splatting/train.py -s data/Bureau_sdf_low_poly/ --iterations 7000 -m outputs_GS/Bureau

Train the Sugar model

python train.py -s data/Bureau_sdf_low_poly/ -c outputs_GS/Bureau/ -r "sdf" --low_poly True

image

image

image

2. Running separate scripts

2.1 extract_mesh.py

    # Mesh computation parameters
    fg_bbox_factor = 1.  # 1.
    bg_bbox_factor = 1.  # 4.
    poisson_depth = 10  # 10
    decimate_mesh = True
    clean_mesh = True
python extract_mesh.py -s data/Bureau_sdf_low_poly_bgfactor_1/ -c outputs_GS/Bureau/ -i 7000 -m output/coarse/Bureau_sdf_low_poly/sugarcoarse_3Dgs7000_sdfestim02_sdfnorm02/15000.pt -l 0.3 -d 200000 -o output/coarse_mesh
image image

2.2 train_refined.py

python train_refined.py -s data/Bureau_sdf_low_poly_bgfactor_1/ -c outputs_GS/Bureau/ -i 7000 -m output/coarse_mesh/Bureau_sdf_low_poly_bgfactor_1/sugarmesh_3Dgs7000_sdfestim02_sdfnorm02_level03_decim200000.ply -o output/refined/Bureau_sdf_low_poly_bgfactor_1/ -v 200000

2.3 extract_refined_mesh_with_texture.py

python extract_refined_mesh_with_texture.py -s data/Bureau_sdf_low_poly_bgfactor_1/ -c outputs_GS/Bureau/ -i 7000 -m output/refined/Bureau_sdf_low_poly_bgfactor_1/sugarfine_3Dgs7000_sdfestim02_sdfnorm02_level03_decim200000_normalconsistency01_gaussperface1/15000.pt -o output/refined_mesh/Bureau_sdf_low_poly_bgfactor_1

image

image