An open API service indexing awesome lists of open source software.

https://github.com/simbleau/spotify-audio-analysis

Deep learning methods for reverse-engineering the Echo Nest Audio Analysis using a large collection of segment frequency spectra and their corresponding pitch, timbre, and loudness as determined by Jehan’s algorithms.
https://github.com/simbleau/spotify-audio-analysis

Last synced: 9 months ago
JSON representation

Deep learning methods for reverse-engineering the Echo Nest Audio Analysis using a large collection of segment frequency spectra and their corresponding pitch, timbre, and loudness as determined by Jehan’s algorithms.

Awesome Lists containing this project

README

          

# Spotify Audio Analysis
Spotify Audio Analysis Repo for CS 5440

# Meeting Room
Team meeting room is **202-A**.

# (Violet) Team Members
- Imbleau, Spencer
- Frazier, Maclean
- Lapenseee-Rankine, Danielle
- Reeves, Andrew
- Lovingood, Jake

# Tools, IDEs, Environments
- Python 3.7+
- Anaconda 3 - [Install](https://docs.anaconda.com/anaconda/install/)
- PyCharm Community Edition - [Install](https://www.jetbrains.com/pycharm/download/)

# Getting started
- After cloning, you **must** put `spotify_train.npz` and `spotify_valid.npz` in the `data/` folder! These are too big for GitHub to track.
- If the conda environment has never been created: ` conda env create -f cs5440.yml`
- Activating the conda environment: `conda activate cs5440`
- Program run example: `python3 loudness.py`
- To check the best models' loss: `python3 best_models.py`

# Troubleshooting
- `ImportError: cannot import name seaborn`
- Fix: `pip install seaborn`
- `"AttributeError: 'str' object has no attribute 'decode' "` , while Loading a Keras Saved Model
- Fix: `pip install 'h5py==2.10.0' --force-reinstall`

# WebCAT
[http://webcatvm.cs.appstate.edu:8080/Web-CAT](http://webcatvm.cs.appstate.edu:8080/Web-CAT)