Near-optimal Active Reconstruction

This repository contains the accompanying code to my bachelor thesis on "Near-optimal Active Reconstruction".

It consists of two interactive notebooks:

• 2D_simulation.ipynb provides an interactive environment for running and visualizing different algorithms.
• 2D_experiments.ipynb provides an interactive environment for evaluating different algorithms by running multiple simulation experiments.

Setup

There are two options for running the interactive notebooks.

For the best performance and compatibility, we recommend the LOCAL setup.

LOCAL (Jupyter Notebook, VS Code with Jupyter Extension)

1. Clone this Git repository and install all requirements.

   git clone https://github.com/danielyxyang/active_reconstruction.git
   cd active_reconstruction
   pip install -r requirements.txt
   

2. Open the folder active_reconstruction in Jupyter or VS Code.
3. Open the desired notebook in the notebooks folder.
4. Set the ENV parameter in the first cell to "LOCAL".
5. Run all cells.

If the notebooks do not work as expected, try out the exact Python package versions.

pip install -r requirements_strict.txt

COLAB (Google Colab)

1. Open the desired notebook from the notebooks folder in Google Colab.
2. Set the ENV parameter in the first cell to "COLAB".
3. Run all cells.

License

This code is released under the MIT license. See LICENSE for details.