What range of tilt angles can frequencies of a drone rotor be identified in spectral analysis of LiDAR pulses on a stationary drone rotor?
This directory contains a collection of miscellaneous scripts that are crucial for specific data processing and model training tasks.
-
calc_norms.py 🧮
- Calculates normalization values for consistent input scaling across the dataset.
-
split_data.py 🔀
- Splits HDF5 files into training, validation, and test sets, ensuring balanced data distribution.
🛠️ Stores custom configuration files for mmdet models, facilitating specific adaptations or optimizations needed for the project.
🖼️ Some outputs from random images from COCO using the mmdet inference demo. Note: These will need to be deleted in the final version.
📚 The cloned repository of the mmdetection library. Includes a bit of customization necessary to get the inference demo running.
💻 Stores scripts for running larger scale jobs on Tempest.
📊 Contains code for generating spectrograms. Generates annotations as well. More details can be found in the specs/README.md file.
🌐 Contains scripts for transferring files using the Globus CLI.
The Docker image is contained here. For use at MSU, the image was converted to an Apptainer image so it could be run on Tempest (MSU HPC).
Note: The Apptainer container (.sif) is not included as the file is too large.
This project includes files from the mmdetection project, which is licensed under the Apache License 2.0. For more information, see the LICENSE-APACHE-2.0 file.