Driverless Tech Pioneer Predicts Multiple Winners In Autonomous Car Race

Standard

Forbes: Can’t the deep-learning approach to training automotive AI solve the localization issue?Iagnemma: “I believe it’s exactly the opposite when we look at a deep-learning based approach. The promise of that approach is that we can teach the car how to drive by feeding it many, many examples of good driving behavior.“If we believe that’s true those examples, if you’re driving in the United States are really not very useful when you’re in Singapore. In fact, they’ll put you on the wrong side of the road.“My point is that when you enter a new market and that market has different rules and a market has different appearances the burden on a learning-based approach may even be higher than an alternative approach because you may have to effectively retrain, if not from scratch, nearly from scratch.”Forbes: What do you consider a better approach?Iagnemma: “Think about a more flexible architecture where you’ve got a more systematic way of enforcing rules of the road.

Source: Driverless Tech Pioneer Predicts Multiple Winners In Autonomous Car Race

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s