“A full build of Autopilot neural networks involves 48 networks that take 70,000 GPU hours to train. Together, they output 1,000 distinct tensors (predictions) at each timestep.”
Tesla Autopilot may be considered a breakthrough, but there are still concerns about semi-autonomous and fully autonomous vehicle technology. The concerns surrounding Tesla stem from the fact that it already has the tech employed in its cars, albeit in beta form. However, due to this real-world testing, Tesla has recorded over 3 billion miles of Autopilot data for use in its research.
The Autopilot section of Tesla’s website explains the technology via several categories, including hardware, neural networks, autonomy algorithms, code foundation, and evaluation infrastructure. We’ll be first to admit that we don’t have a solid grasp of AI and neural networks, but the information is definitely fascinating. Even more interesting is the latest video, which is embedded in the tweet below.
The above video is also shown on Tesla’s website, along with the section on neural networks. According to Tesla:
“Our per-camera networks analyze raw images to perform semantic segmentation, object detection and monocular depth estimation. Our birds-eye-view networks take video from all cameras to output the road layout, static infrastructure and 3D objects directly in the top-down view. Our networks learn from the most complicated and diverse scenarios in the world, iteratively sourced from our fleet of nearly 1M vehicles in real time. A full build of Autopilot neural networks involves 48 networks that take 70,000 GPU hours to train. Together, they output 1,000 distinct tensors (predictions) at each timestep.”
We’d love to know your thoughts on Tesla Autopilot. Let us know in the comment section below.