DeepPiCar — Part 6: Traffic Sign and Pedestrian Detection and Handling
<p>In <a href="https://towardsdatascience.com/deeppicar-part-4-lane-following-via-opencv-737dd9e47c96" rel="noopener" target="_blank">Part 4</a> and <a href="https://medium.com/@dctian/deeppicar-part-5-lane-following-via-deep-learning-d93acdce6110" rel="noopener">Part 5</a> of the blog series, we discussed lane detection and navigation. A true autonomous vehicle would also need to be aware of its surroundings at all times. In this article, we will discuss another important perception feature, namely, <strong>detecting traffic signs and pedestrians</strong>. Note this feature is not available in any 2019 vehicles, except <a href="https://electrek.co/2019/03/28/tesla-autopilot-stopping-red-light/" rel="noopener ugc nofollow" target="_blank">maybe Tesla</a>. We will train DeepPiCar to <strong>identify </strong>and <strong>respond </strong>to (miniaturized) traffic signs and pedestrians <strong>in real time</strong>. We will also utilize the $75 Google Edge TPU accelerator that just came out in early 2019!</p>
<p>Note to readers who have not purchased any hardware: For the first half of this article, you can follow along with just a PC, since we will be training our model in the Cloud.</p>
<p><a href="https://towardsdatascience.com/deeppicar-part-6-963334b2abe0"><strong>Click Here</strong></a></p>