Installation

# Clone the repository
git clone https://github.com/christoaluckal/gaussian-splatting.git --recursive
cd gaussian-splatting

# Create and activate environment (recommended)
conda env create --file environment.yml
conda activate gaussian_splatting

Running the Code

# Run the partitioning code from a COLMAP txt data source. This will create 'output folder/model0' and 'output folder/model1'
python colmap_splitter/split.py -s 'original_COLMAP_txt_scene' -m 'output_folder' -f 'frame_name' --num_test 'number_of_test_images'

# Run 3D Gaussian Splatting on the split dataset
python train.py \
  -s output_folder/model0 \        # Output folder
  -m output_name \                 # Model name
  -r resolution_fraction \         # Downscale resolution factor (e.g., 2 for half-res)
  --eval \                         # Use test images for evaluation
  --pkl_name output_pkl_name \     # Save metrics (time, loss, PSNR, etc.) to this PKL
  -x 1 \                           # Number of extensions (1 when using split.py)
  --splitter_itr iteration_value   # Iteration frequency for adding new Gaussian subsets

Output

The trained Gaussian model and reconstruction results will be saved to the output/ directory.