### Setup SplaTAM Docker Environment Source: https://github.com/spla-tam/splatam/blob/main/README.md Sets up and runs SplaTAM within a Docker container. It pulls a pre-built Docker image, starts the container, activates a virtual environment inside, and installs dependencies. ```bash docker pull nkeetha/splatam:v1 bash bash_scripts/start_docker.bash cd /SplaTAM/ pip install virtualenv --user mkdir venv cd venv virtualenv --system-site-packages splatam source ./splatam/bin/activate pip install -r venv_requirements.txt ``` -------------------------------- ### Setup SplaTAM Singularity Container Source: https://github.com/spla-tam/splatam/blob/main/README.md Configures and launches SplaTAM using a Singularity container. This process involves pulling a Singularity image, starting an instance, and setting up a virtual environment for dependencies. ```bash cd singularity pull splatam.sif docker://nkeetha/splatam:v1 singularity instance start --nv splatam.sif splatam singularity run --nv instance://splatam cd pip install virtualenv --user mkdir venv cd venv virtualenv --system-site-packages splatam source ./splatam/bin/activate pip install -r venv_requirements.txt ``` -------------------------------- ### Run SplaTAM Reconstruction Source: https://github.com/spla-tam/splatam/blob/main/README.md Executes the SplaTAM algorithm to perform 3D reconstruction using a specified configuration file. This is the primary command to start the reconstruction process. ```bash python scripts/splatam.py configs/iphone/splatam.py ``` ```bash python scripts/splatam.py configs/replica/splatam.py ``` ```bash python scripts/splatam.py configs/replica/splatam_s.py ``` ```bash python scripts/splatam.py configs/tum/splatam.py ``` -------------------------------- ### Install SplaTAM with Conda and Pip Source: https://github.com/spla-tam/splatam/blob/main/README.md Installs SplaTAM and its dependencies using Conda for environment management and CUDA toolkit, followed by Pip for project-specific requirements. Assumes Python 3.10, Torch 1.12.1, and CUDA 11.6. ```bash conda create -n splatam python=3.10 conda activate splatam conda install -c "nvidia/label/cuda-11.6.0" cuda-toolkit conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.6 -c pytorch -c conda-forge pip install -r requirements.txt ``` -------------------------------- ### Download Replica Dataset Source: https://github.com/spla-tam/splatam/blob/main/README.md Downloads the Replica dataset using a bash script. The data will be saved to the default `./data/Replica` directory. This is a prerequisite for running SplaTAM on Replica data. ```bash bash bash_scripts/download_replica.sh ``` -------------------------------- ### Run Online Demo with iPhone Capture Source: https://github.com/spla-tam/splatam/blob/main/README.md Executes the SplaTAM online demo using data streamed from an iPhone via the NeRFCapture app. Requires the iPhone and PC to be on the same WiFi network. The script launches the demo and allows interactive rendering on the PC. ```bash bash bash_scripts/online_demo.bash configs/iphone/online_demo.py ``` -------------------------------- ### ScanNet Data Processing (Conceptual) Source: https://github.com/spla-tam/splatam/blob/main/README.md This entry represents the conceptual process of obtaining and preparing ScanNet data. It involves downloading from the official website and using a Python script to extract frames from .sens files. The exact download command is not provided but is described as a procedure. -------------------------------- ### Visualize Final SplaTAM Reconstruction Source: https://github.com/spla-tam/splatam/blob/main/README.md Renders the final, interactive SplaTAM reconstruction using a given configuration file. This command allows for detailed inspection of the generated 3D model. ```bash python viz_scripts/final_recon.py configs/iphone/splatam.py ``` -------------------------------- ### Download TUM-RGBD Dataset Source: https://github.com/spla-tam/splatam/blob/main/README.md Downloads the TUM-RGBD dataset using a bash script. This command is used to acquire the necessary data for running SplaTAM on TUM-RGBD sequences. ```bash bash bash_scripts/download_tum.sh ``` -------------------------------- ### Visualize SplaTAM Reconstruction Online Source: https://github.com/spla-tam/splatam/blob/main/README.md Enables real-time visualization of the SplaTAM reconstruction as it is being generated. This is useful for monitoring the process and making adjustments on the fly. ```bash python viz_scripts/online_recon.py configs/iphone/splatam.py ``` -------------------------------- ### Run 3D Gaussian Splatting with Ground Truth Poses Source: https://github.com/spla-tam/splatam/blob/main/README.md Executes the 3D Gaussian Splatting algorithm using provided ground truth poses for a dataset. This is typically used for benchmarking or when accurate pose information is available. ```bash python scripts/gaussian_splatting.py configs/iphone/gaussian_splatting.py ``` -------------------------------- ### Collect iPhone Dataset for SplaTAM Source: https://github.com/spla-tam/splatam/blob/main/README.md Initiates the dataset collection process using the NeRFCapture app on an iPhone. This command is used solely for capturing data and preparing it for offline processing with SplaTAM. ```bash bash bash_scripts/nerfcapture2dataset.bash configs/iphone/dataset.py ``` -------------------------------- ### Post-Splatting Optimization for SplaTAM Source: https://github.com/spla-tam/splatam/blob/main/README.md Runs a post-processing optimization step on the SplaTAM reconstruction to refine the 3D Gaussian splats. This command improves the quality and appearance of the final output. ```bash python scripts/post_splatam_opt.py configs/iphone/post_splatam_opt.py ``` -------------------------------- ### Run SplaTAM Offline Demo on Captured Dataset Source: https://github.com/spla-tam/splatam/blob/main/README.md Processes a pre-captured dataset with SplaTAM in offline mode. This command assumes the dataset has already been collected using the NeRFCapture app. ```bash bash bash_scripts/nerfcapture.bash configs/iphone/nerfcapture.py ``` -------------------------------- ### Export SplaTAM Splats to PLY Source: https://github.com/spla-tam/splatam/blob/main/README.md Exports the generated splats from the SplaTAM reconstruction into a .ply file format. This allows for compatibility with various 3D viewers and further processing. ```bash python scripts/export_ply.py configs/iphone/splatam.py ``` === COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.