TwilightSLAM project created at the University of Michigan Ann Arbor with the objective of investigating state-of-the-art technologies for Simultaneous Localization and Mapping (SLAM) tasks in low-light or twilight-like environments.

As this project operates in correspondence with various libraries and core algorithmic frameworks (keeping in mind quick real-time implementation ability), system requirements vary for each repository. Each image enhancment module and SLAM framework has a detailed readme associated in the corresponding directory for smooth implementation of our combined Twilight SLAM architectures.

For a brief overview of our project, we recommend watching our presentation on YouTube. For a more indepth look at our project and the results, you can find the PDF of our report here or on arxiv

Each type of SLAM implementation has a detailed readme associated in the corresponding directory for smooth implementation of our combined Twilight SLAM architectures. Go Blue!

P.S. this is a combined and updated implementation with recent python and C++ library changes to our original project for concise viewing and version-based bug free experience