01. Project Intro

That was a lot of information to take in, but now it's time to put it all into practice! In this project, you will build your own segmentation network, train it, validate it, and deploy it in the Follow Me project. In the end, you will have your own drone tracking and following a single hero target using your own network pipeline!

If you have completed the Lab: Semantic Segmentation, you are in a good position to tackle the Follow Me project. If not, you could follow the steps in the last lesson and give yourself a boost for the final project!

Project Steps:

1. Setting up your local environment: This is how you will test your initial network design to make sure there are no errors. Otherwise, you can use up your Workspace GPU hours or rack up charges on Amazon Web Services (AWS) with simple debugging instead of training. You will also use your local environment to evaluate your Workspace/AWS trained model with the Follow Me simulator.

2. Brief overview on how the simulator works and its basic controls.

3. Collecting data from the simulator to train your network.

4. Building your neural network.

5. Setting up your Classroom Workspace or, if you prefer, setting up your AWS Amazon Machine Images (AMI) .

6. Training your network and extracting your final model and weights from Udacity GPU Workspace or AWS Instance.

7. Downloading your model from cloud and testing your model with the Follow Me simulator.

8. Getting your project ready for the final submission!