A TensorFlow implementation of this Nvidia paper with some changes.
- NVidia dataset: 72 hrs of video => 726060*30 = 7,776,000 images
- Nvidia blog: https://devblogs.nvidia.com/deep-learning-self-driving-cars/
- Research paper: End to End Learning for Self-Driving Cars by Nvidia. [https://arxiv.org/pdf/1604.07316.pdf]
- More data: https://medium.com/udacity/open-sourcing-223gb-of-mountain-view-driving-data-f6b5593fbfa5
- https://medium.com/udacity/open-sourcing-223gb-of-mountain-view-driving-data-f6b5593fbfa5
- https://aws.amazon.com/blogs/machine-learning/get-started-with-deep-learning-using-the-aws-deep-learning-ami/
- https://www.youtube.com/watch?v=qhUvQiKec2U
- https://medium.com/udacity/teaching-a-machine-to-steer-a-car-d73217f2492c
You need to have installed following softwares and libraries before running this project.
- Python 3: https://www.python.org/downloads/
- Anaconda: It will install ipython notebook and most of the libraries which are needed like sklearn, pandas, seaborn, matplotlib, numpy and scipy: https://www.anaconda.com/download/
Download the dataset and extract into the repository folder
Use python train.py to train the model
Use python run.py to run the model on a live webcam feed
Use python run_dataset.py to run the model on the dataset
To visualize training using Tensorboard use tensorboard --logdir=./logs, then open http://0.0.0.0:6006/ into your web browser.
-
tensorflow: TensorFlow provides multiple APIs.The lowest level API, TensorFlow Core provides you with complete programming control.
- pip install tensorflow
- conda install -c anaconda tensorflow
-
opencv: OpenCV-Python is the Python API of OpenCV. It combines the best qualities of OpenCV C++ API and Python language.
- conda install -c conda-forge opencv