Downloads the Udacity Simulator dataset and loads it as pandas dataframes. Images are not loaded into memory but the center, left, and right column contain the relative path to their files.

import dataget

df = dataget.image.udacity_simulator().get()

Do not use random shuffling to spit the data as similar images from the training set will appear on the test set.




type shape
df pd.DataFrame (8036, 7)


column type description
center str relative path of the center camera jpg image
left str relative path of the left camera jpg image
right str relative path of the right camera jpg image
steering float64 stearing angle
throttle float64 throttle magnitude
brake float64 break mangnitude
speed float64 speed mangnitude


  • Folder name: image_udacity_simulator
  • Size on disk: 465MB