Stereo Frustums: A Siamese Pipeline for 3D Object Detection
This work addresses 3D object detection for autonomous vehicles, presenting an incremental improvement over existing stereo-based methods.
The paper tackles 3D object detection for autonomous driving by proposing a stereo frustums matching module that uses 2D proposals and epipolar constraints to avoid traditional stereo matching, achieving state-of-the-art results on the KITTI dataset.
The paper proposes a light-weighted stereo frustums matching module for 3D objection detection. The proposed framework takes advantage of a high-performance 2D detector and a point cloud segmentation network to regress 3D bounding boxes for autonomous driving vehicles. Instead of performing traditional stereo matching to compute disparities, the module directly takes the 2D proposals from both the left and the right views as input. Based on the epipolar constraints recovered from the well-calibrated stereo cameras, we propose four matching algorithms to search for the best match for each proposal between the stereo image pairs. Each matching pair proposes a segmentation of the scene which is then fed into a 3D bounding box regression network. Results of extensive experiments on KITTI dataset demonstrate that the proposed Siamese pipeline outperforms the state-of-the-art stereo-based 3D bounding box regression methods.