ChallengeRocket

NVIDIA® Jetson™ Developer Challenge

CURRENT STATUS

Winners announced

SEE RESULTS

Thanks to all Participants!

SEE RESULTS

NVIDIA® Jetson™ Developer Challenge

Back

rAInger

Project description

rAInger is a piece of software that transforms an Nvidia Jetson into an autonomous wildlife officer. It uses a connected camera to record images, leverages the GPU and a deep learning network to analyze them, and transmits metadata from the analysis over a low-power high range antenna to some central database. This metadata could include

  1. What kinds of animals it has seen
  2. Where it has seen them, and when
  3. The location of some forest disturbance (a wildfire, for example)
  4. The location of any wireless signals in the forest (in case someone was lost, for example)


The first three can be implemented using a convolutional neural network trained to recognize various forest occurrences (animals, emergencies, etc.)  The latter is made possible through the LoRaWAN antenna, which uses an efficient MAC protocol already being leveraged here in Boston

Having this system in the wild can give us accurate species distribution data throughout an area, assist emergency responders with precise location information, and provide a reliable emergency network throughout the country should more traditional means fail.

Wildlife cameras are also a fun way for the public to engage with our local fauna, and by making them smarter we may make them more accessible to the public and increase wildlife awareness.

Please see the video for more information, and thank you for your time.