Deep learning to support native birds

Dennis Sosnoski is a deep learning engineer working with The Cacophony
Project (https://cacophony.org.nz/) on developing models to
automatically classify sounds and images. The goal of The Cacophony
Project is to help eliminate threats to our native birds and encourage
the recovery of their populations. An important part of the work being
done by Cacophony is monitoring and trapping of introduced predator
species, and as part of this they're sourcing infrared video cameras to
individuals and organizations to observe predator activity in the
environment (such as
https://www.stuff.co.nz/environment/300288666/thermal-camera-captures-stoats-at-aucklands-pestfree-shakespear-regional-park).
Previously the recordings from these cameras had to be viewed and tagged by a human for accurate identification of animals caught on-camera, but new models based on deep learning are allowing Cacophony to switch to automated tagging of the videos. This makes it practical to deploy the cameras to monitor more territory at lower cost., bringing us all a big step closer to quantifying the predator problem in Aotearoa and finding the best solutions.

In this talk Dennis will give a background on how the infrared camera
videos are being processed by Cacophony, discuss the work the Cacophony team has done on classifying videos, and demonstrate the operation of the current best model on recorded videos. He'll also walk us through an overview of the model architectures and techniques used for training the models, including showing how data representation has played an important role in improving the accuracy of the models.

If you're interested in the details of the models and training you may want to attend Dennis' follow-up meeting on Thursday evening for a technical discussion of the deep learning work using TensorFlow: https://www.meetup.com/wellington-tensorflow-meetup-group/events/279143678/