Bundle-Adjusting NeuS

Mohamed S. Sabae · Hoda A. Baraka · Mayada M. Hadhoud

Original Paper | Extended Paper | Project Page

Neural implicit surface reconstruction method that extends NeuS by enabling joint camera parameters optimization.

Installation

• Clone the repository:

git clone https://github.com/DarkGeekMS/bundle-adjusting-neus.git
cd bundle-adjusting-neus
pip install -r requirements.txt

• Refer to NeuS for more information on data convention.

Usage

• Optimize with no pose priors:

python main.py --mode train --conf ./configs/ba_no_poses.conf --case <case_name>

• Optimize with noisy pose priors:

python main.py --mode train --conf ./configs/ba_noisy_poses.conf --case <case_name>

• Optimize with no pose priors (with camera intrinsics):

python main.py --mode train --conf ./configs/ba_no_poses_int.conf --case <case_name>

• Optimize with noisy pose priors (with camera intrinsics):

python main.py --mode train --conf ./configs/ba_noisy_poses_int.conf --case <case_name>

• Extract surface from trained model:

python main.py --mode validate_mesh --conf <config_file> --case <case_name> --is_continue # use latest checkpoint

The corresponding mesh can be found in exp/<case_name>/<exp_name>/meshes/<iter_steps>.ply.

• View interpolation:

python main.py --mode interpolate_<img_idx_0>_<img_idx_1> --conf <config_file> --case <case_name> --is_continue # use latest checkpoint

The corresponding image set of view interpolation can be found in exp/<case_name>/<exp_name>/render/.