A semi-synthetic approach to creating visual-inertial odometry datasets.
University of Canterbury
Time & Place
Thu, 03 Jun 2021 12:00:00 NZST in Rehua 529
Capturing outdoor visual-inertial datasets is a challenging yet vital aspect of developing robust visual-inertial odometry (VIO) algorithms. A significant hurdle is that high-accuracy-ground-truth systems (e.g., motion capture) are not practical for outdoor use. One solution is to use a ``semi-synthetic" approach that combines rendered images with real IMU data. This approach can produce sequences containing challenging imagery and accurate ground truth but with less simulated data than a fully synthetic sequence. Existing methods record IMU measurements from a visual-inertial system while measuring its trajectory using motion capture, then rendering images along that trajectory. We propose fusing motion capture and IMU measurements to generate a more accurate trajectory for the rendering process. We demonstrate that this improved trajectory results in better consistency between the IMU data and rendered images. Furthermore, we show that VIO trajectory error is reduced by 79% when evaluated on sequences generated using our approach compared to existing methods.
Sam is a PhD student focusing on improving the robustness of visual odometry in challenging environments. He is also a research assistant for UCDroneLab working on autonomous UAVs.