To get started with the dataset, see the README file on the dataset’s GitHub repository.
We provide software
templates to access the
dataset. If you are not familiar with ROS and
rosbag in particular, we
recommend you to start from here.
Tools to analyse the dataset are split into two:
basic scripts are available to generate quick statistics on the dataset (lengths of recordings, demographics…) and manipulate the dataset files (check integrity, copy, etc.). These scripts can be found in the /tools directory of the dataset meta-data repository.
more advanced analysis tools that typically require compilation are also available.
More documentation to be added soon – if you have specific questions/needs, please open a ticket here.
The reference coding scheme, as well as our custom tool to annotate recordings, is explained here.