Closed arichadda closed 1 year ago
Base: 84.02% // Head: 84.02% // No change to project coverage :thumbsup:
Coverage data is based on head (
84d576e
) compared to base (8c6b5ef
). Patch has no changes to coverable lines.
:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
This PR adds scripts for performing k-Means clustering on data collected by the AI Sonobuoy and creating Uniform Manifold Approximation and Projection (UMAP) visualizations on several canonical 1-dimensional audio features.
Note: includes support for NVIDIA's RAPIDS
cuML
library1_Dimensional_Audio_Features.ipynb
: Creates several canonical 1-D audio features based on the collected data to cluster including: amplitude envelope, root mean squared energy, zero crossing rate, spectral flux, spectral bandwidth, spectral centroid, and band energy ratio.clustering.py
: performs k-Means clustering and tunes the "number of neighbor" hyperparameter to the trustworthiness metric for UMAP as well as several displays of summary metrics, etc.