Development of bioacoustic analytical tools for both humans (citizen/scientists) and machines (algorithms) to process Orcasound data -- either post-processing of archived raw FLAC files or real-time analyis of the lossy stream and/or FLAC files. The long-term goal is to promote a friendly competition between humans and machines that leads to synergistic real-time, cloud-based processing of acoustic data.

Resources:

• Data for Good tarball of sound samples

Resources to develop:

Archive of signals, noise, and empirical data for machine learning and teaching human listeners

• Example of Orcasound FLAC files (48, 96, 192 kHz)
• Guidance on how to access S3 buckets (CLI and/or Cloud9)

Overview of AWS tools and plan(s) for utilizing them

• EC2
• Cloud9
• Lamba
• Batch
• ECS

Experiments in cloud-based bio/acoustic analysis

• How to get up to speed with Cloud9 IDE and access to Orcasound data in S3 buckets
• Val's initial Python scripts

Other related open-source projects, and tools for testing tools (e.g. algorithms) with Orcasound data

• Demonstrate how to run a Pamguard module on Orcasound data (archived first; then real-time)
• Ishmael?
• Triton?