03. Classification and Localization

Distracted Driver Detection

As was mentioned in the video, detecting distracted drivers is a classification and localization challenge. As part of this task, some approaches localize cell phones in a image; if a cell phone was detected as close to a driver's face, then it assumes that the driver is on the phone and distracted, but if a phone is located on a car seat or far away from the driver, then the driver is more likely to be focused on the road. Check out an example of such detection code in Keras.