How many of us would consider turning our passion for the arts into something that would help a species from endangerment? Meet Daniel DeLeon, a 24-year-old passionate musician who has, through machine learning, using TensorFlow, developed an instrument that can track and record whale calls (Fin and Blue) and their changing migration patterns to study the broader effects of human influence on marine life.
Daniel teamed up with John Ryan and Danelle Cline at the Monterey Bay Aquarium Research Institute (MBARI), where he applied for a summer internship. They were already using machine learning to study the sounds of whales. This sparked Daniel's curiosity, and he wanted to differentiate between the calls of Blue whale and Fin whale to understand the ecology better.
The MBARI already has accumulated thousands of audio files from the ocean; segregating it would be a strenuous task. This is where machine learning came into play. ML enabled them to go through these files and track down the particular sounds that they trained the model to identify using TensorFlow. “This tracking system can tell us how many calls were made in any given amount of time near the Monterey Bay, and will enable scientists at MBARI to track their changing migration behavior, and advance their research on whale ecology and how human influence above water negatively impacts marine life below,” said Daniel to Google, who owns TensorFlow.
Blue and Fin whales are some of the loudest animals on earth. Their low-frequency calls can travel across long stretches of ocean, making them excellent candidates for study. MBARI’s hydrophone can hear whales up to 500 kilometers away.
The sound waves recorded by the hydrophone are converted into visual data in the form of spectrograms, a map of sound waves that is then fed into the TensorFlow that learns through repeated training what Blue and Fin whales look like. Daniel feeds the TensorFlow with over 18,000 samples, for a 98 per cent accuracy. Now it can differentiate between the whales, tell us what time the call was made and how loud it was, and how long it lasted.
“We’re at a very pivotal point in ocean science. It’s a really interesting time for machine learning too, because we are Finally starting to solve problems that we weren’t able to even five years ago.” Said Danelle Cline, to Google.
Daniel’s work with machine learning has enabled Danelle and John to focus on more important things—change in migration patterns that have been followed for eons, the broader impact on marine life below water by human above water due to pollution and climatic change.