HS-Pose: Hybrid Scope Feature Extraction for Category-level Object Pose Estimation
This work addresses the problem of accurate pose estimation for objects with large shape variations, which is crucial for robotics and augmented reality applications, and represents an incremental improvement over existing methods.
The paper tackles category-level object pose estimation by proposing a hybrid scope feature extraction layer (HS-layer) to overcome limitations of 3D graph convolution methods, achieving a 14.5% improvement on 5d2cm and 10.3% on IoU75 over the baseline and outperforming state-of-the-art by 8.3% and 6.9% on the REAL275 dataset.
In this paper, we focus on the problem of category-level object pose estimation, which is challenging due to the large intra-category shape variation. 3D graph convolution (3D-GC) based methods have been widely used to extract local geometric features, but they have limitations for complex shaped objects and are sensitive to noise. Moreover, the scale and translation invariant properties of 3D-GC restrict the perception of an object's size and translation information. In this paper, we propose a simple network structure, the HS-layer, which extends 3D-GC to extract hybrid scope latent features from point cloud data for category-level object pose estimation tasks. The proposed HS-layer: 1) is able to perceive local-global geometric structure and global information, 2) is robust to noise, and 3) can encode size and translation information. Our experiments show that the simple replacement of the 3D-GC layer with the proposed HS-layer on the baseline method (GPV-Pose) achieves a significant improvement, with the performance increased by 14.5% on 5d2cm metric and 10.3% on IoU75. Our method outperforms the state-of-the-art methods by a large margin (8.3% on 5d2cm, 6.9% on IoU75) on the REAL275 dataset and runs in real-time (50 FPS).