Table of Contents

• Description
• Public Domain Annotations
• Citations and Repositories
• Notation
• Installation
  • Downloading Executables
  • System Configuration
  • Scheduling Jobs
    • Locally
    • Another Workstation
    • An On-Premise Cluster
• Starting SongExplorer
• Tutorial
  • Manually Annotating
  • Training a Classifier
  • Quantifying Accuracy
  • Making Predictions
  • Fixing Mistakes
  • Minimizing Annotation Effort
• Advanced Usage
  • Detecting Rare Sounds
  • Clustering Annotations
  • Correcting False Alarms
  • Correcting Misses
  • Double Checking Annotations
  • Measuring Generalization
  • Searching Hyperparameters
  • Limiting Ground Truth
  • Examining Errors
  • Ensemble Models
  • Transfer Learning
  • Autoencoding
  • Testing Densely
  • Discovering Novel Sounds
  • Overlapped Classes
• Unsupervised Methods
• Scripting Automation
• Training on Video
• Customizing with Plug-ins
  • Loading Data
  • Video Filenames
  • Event Detection
  • Double-Click Annotations
  • Network Architecture
  • Clustering Algorithm
  • Augmentation
• Troubleshooting
• Frequently Asked Questions
• Reporting Problems
• Development
  • Conda
  • Singularity
  • Docker
  • System Tests

Description

You have an audio recording, and you want to know where certain classes of sounds are. SongExplorer is trained to recognize such words by manually giving it a few examples. It will then automatically calculate the probability, over time, of when those words occur in all of your recordings.

Applications suitable for SongExplorer include quantifying the rate or pattern of words emitted by a particular species, distinguishing a recording of one species from another, and discerning whether individuals of the same species produce different song.

Underneath the hood is a deep convolutional neural network. The input is the raw audio stream, and the output is a set of probability waveforms corresponding to each word of interest.

Training begins by first manually annotating a few sounds with however many word labels naturally occur. A classifier is then trained on this corpus of ground truth, and a new recording is analyzed by it. The words it automatically finds are then displayed with predicted labels. You manually correct the mistakes, both re-labeling words that it got wrong, as well as labeling words it missed. These new annotations are added to the ground truth, and the process of retraining the classifier and analyzing and correcting new recordings is repeated until the desired accuracy is reached.

Public Domain Annotations

SongExplorer is open source and free for you to use. However, SongExplorer is not a static piece of software. It’s performance is improved with additional high-quality annotations.

Therefore, when you publish results based on SongExplorer, we request that you make all of your primary data and annotations freely available in a recognized data repository, such as figshare, Dryad, or Zenodo. Many journals already require deposition of raw data, but we strongly encourage you to also provide your manual annotations. These manual annotations will serve to improve the performance of SongExplorer over time, helping both your own work and that of everyone else.

Please let us know where you have