HCII Paper and Presentation
Released today, I have a new published paper and presentation! Together with two of my work colleagues, I published a paper at Human Computer Interaction International 2021 (HCII). The paper and associated talk presents a tool we created for generating labeled synthetic data for neural network training. The novelty of the tool is that it is written entirely within the ROS architecture. The paper, source code, and presentation can all be found below.
The purpose of the tool is to assist the user in creating perfectly labeled synthetic data using ROS and Gazebo. While there are other such tools, none leverage the ROS ecosystem. My team and I wanted to create something that would be extremely familiar to someone already using ROS, as opposed to something like Unity that would require additional learning. The tool can also write data to multiple formats simultaneously. For example, it can write to COCO’s JSON format and YOLO’s text file format. It can also easily be extended to include additional formats as required. This is useful when you are exploring different networks that may require data in different formats.
The tool should provide an easy user experience to anyone looking to generate labeled data automatically via ROS. The code is freely available at GitHub and I encourage anyone to check it out if you think you will have a use for it. And since it is open source, feel free to make any changes, note any bugs, or suggest improvements. I would love to hear from you!