Expansion of Visual Hints for Improved Generalization in Stereo Matching
This work addresses generalization issues in stereo matching for computer vision and robotics applications, though it appears incremental as it builds on existing visual odometry techniques.
The paper tackles the problem of improving generalization in stereo matching by expanding sparse 3D visual hints using a 3D random geometric graph, resulting in enhanced performance on multiple benchmarks without requiring additional sensors.
We introduce visual hints expansion for guiding stereo matching to improve generalization. Our work is motivated by the robustness of Visual Inertial Odometry (VIO) in computer vision and robotics, where a sparse and unevenly distributed set of feature points characterizes a scene. To improve stereo matching, we propose to elevate 2D hints to 3D points. These sparse and unevenly distributed 3D visual hints are expanded using a 3D random geometric graph, which enhances the learning and inference process. We evaluate our proposal on multiple widely adopted benchmarks and show improved performance without access to additional sensors other than the image sequence. To highlight practical applicability and symbiosis with visual odometry, we demonstrate how our methods run on embedded hardware.