This repo contains the code for demos, exercises, and exercise solutions.
This repository organizes the code by the lessons that they are used in. Each set of code is located in their respective lessons, except for the primary basic_loop.py file that can run each exercise.
Please note that certain instructions for each exercise are only provided within the Udacity classroom.
All lesson 1 files are in /lesson-1-lidar-sensor/.
This directory contains: examples, exercises/starter, and exercises/solution.
Udacity students can make use of the pre-configured workspace environment within the classroom. Alternatively, you can create an environment using the requirements.txt file included in this repository, using a command like pip install -r requirements.txt if you have pip installed, or creating an Anaconda environment in similar fashion.
The Waymo Open Dataset Reader is a very convenient toolbox that allows you to access sequences from the Waymo Open Dataset without the need of installing all of the heavy-weight dependencies that come along with the official toolbox. The installation instructions can be found in tools/waymo_reader/README.md.
This course makes use of three different sequences to illustrate the concepts of object detection and tracking. These are:
- Sequence 1 :
training_segment-1005081002024129653_5313_150_5333_150_with_camera_labels.tfrecord - Sequence 2 :
training_segment-10072231702153043603_5725_000_5745_000_with_camera_labels.tfrecord - Sequence 3 :
training_segment-10963653239323173269_1924_000_1944_000_with_camera_labels.tfrecord
To download these files, you will have to register with Waymo Open Dataset first: Open Dataset – Waymo, if you have not already, making sure to note "Udacity" as your institution.
Once you have done so, please click here to access the Google Cloud Container that holds all the sequences. Once you have been cleared for access by Waymo (which might take up to 48 hours), you can download the individual sequences.
The sequences listed above can be found in the folder "training". Please download them and put the tfrecord-files into the dataset folder within the repository.
It's important to note that certain examples in this lesson make use of the outputs of other examples.
For instance, Example C2-3-3 (and Exercise C2-3-2 before it) use the output of Example C2-3-1.
Otherwise, it is often preferred to comment back out previous examples/exercises so you don't have to cycle through all of the related visualizations.
To traverse frames, you may need to use the Escape key to progress frames. In limited instances, closing a given frame visualization by clicking on the X of the window may also traverse frames.
In Example C2-4-3, add True as an additional argument to render_bb_over_bev() for visualization, but remove it when using Example C2-4-4.