Deep learning based annotation tools extensively adopted by the automotive industry

Deep learning based annotation tools extensively adopted by the automotive industry

Deep learning based annotation tools extensively adopted by the automotive industry

Overview

There are various scenarios on road with different curvatures, turns, slopes, bridges, flyovers and others along with different obstructions
A human driver analyzes the situation to maneuver the vehicle
The algorithm is built to teach a vehicle to drive on its own by cloning the actions a human driver executes to steer the car along various scenarios
Here a CNN (Convolutional Neural Network) algorithm maps the actual data from the steering wheel (as the driver drives the vehicle) with the images for road curvature (from the cameras). Then with number of iterations, CNN learns on its own to build a stronger algorithm that can achieve the required precision and accuracy.
This algorithm clones the human behavior to drive the vehicle on its own

Phase 1 – Training phase

  1. The actual steering data is compared with the road curvature data captured from images using cameras to train the algorithm and increase its accuracy & precision
  2. Images from Camera are used to understand the curvature of road
    CNN (CNN – Convolutional Neural Network) – Understands corresponding steering angle for curvature of road
  3. Actual Rotation & Shift angle from Steering wheel
    CNN (CNN – Convolutional Neural Network) – Understands corresponding steering angle for curvature of road

Phase 2 – Deployment phase

  1. The CNN built in training phase is deployed to generate the steering rotation angle using the data from camera
    CNN (CNN – Convolutional Neural Network)
  2. Network computed steering command to adjust rotation shift & steering angle.