Main Content

Deep Picar

Autonomous cars have been a topic of increasing interest in recent years as many companies are actively developing related hardware and software technologies toward fully autonomous driving capability with no human intervention.

Deep neural networks (DNNs) have been successfully applied in various perception and control tasks in recent years.

They are important workloads for autonomous vehicles as well.

What is a Deep PiCar ?

Deep Picar is a small scale replication of a real self-driving car called DAVE-2 by NVIDIA. Using Deep Picar, we analyze the Raspberry Pi computing capabilities to support end-to-end deep learning based real-time control of autonomous vehicles.

Project Idea :

Controlling of the vehicles by switching from human-led driving to automated self-driving.

PiCar-S Raspberry Pi Smart Robot Car Kit
Raspberry Pi 3 ( Model B+ )
Google Edge TPU coprocessor ( Accelerator )
Micro SD card 16GB
Wide Angle USB Camera 1.3 megapixel with 170 degree fisheye Lens
Adapter HDMI 90 Degree Angle
Battery 18650 ( 3.7V )
Battery charger
Traffic signs”

Link to article