OrcVIO: Object residual constrained Visual-Inertial Odometry
This work addresses the need for more robust and meaningful SLAM systems for robotics and autonomous navigation, though it appears incremental by building on existing object-aware methods.
The paper tackles the problem of improving visual-inertial odometry by integrating object-level semantic information, resulting in accurate trajectory estimation and large-scale object-level mapping as demonstrated on real data.
Introducing object-level semantic information into simultaneous localization and mapping (SLAM) system is critical. It not only improves the performance but also enables tasks specified in terms of meaningful objects. This work presents OrcVIO, for visual-inertial odometry tightly coupled with tracking and optimization over structured object models. OrcVIO differentiates through semantic feature and bounding-box reprojection errors to perform batch optimization over the pose and shape of objects. The estimated object states aid in real-time incremental optimization over the IMU-camera states. The ability of OrcVIO for accurate trajectory estimation and large-scale object-level mapping is evaluated using real data.