In AirsimE2EDeeplearning, we use the trained model to predict the steering angle for car-controls, should the predition angle (the model's output)-multiply 0.69 ?
Problem details
In TrainModel.ipynb,I have noticed that the steering angle is described in draw_image_with_label(img, label, prediction=None) function by this:
theta = label * 0.69 #Steering range for the car is +- 40 degrees -> 0.69 radians
Problem description
In AirsimE2EDeeplearning, we use the trained model to predict the steering angle for car-controls, should the predition angle (the model's output)-multiply 0.69 ?
Problem details
In TrainModel.ipynb,I have noticed that the steering angle is described in draw_image_with_label(img, label, prediction=None) function by this:
theta = label * 0.69 #Steering range for the car is +- 40 degrees -> 0.69 radians
the prediction :
theta = prediction * 0.69
so ,my question is :In TestModel.ipynb,
car_controls.steering = round( float(model_output[0][0]), 2) print('Sending steering = {0}, throttle = {1}'.format(car_controls.steering, car_controls.throttle))
the model's output should multiply 0.69?