CenterSnap: Single-Shot Multi-Object 3D Shape Reconstruction and Categorical 6D Pose and Size Estimation

3 Mar 2022  ·  Muhammad Zubair Irshad, Thomas Kollar, Michael Laskey, Kevin Stone, Zsolt Kira ·

This paper studies the complex task of simultaneous multi-object 3D reconstruction, 6D pose and size estimation from a single-view RGB-D observation. In contrast to instance-level pose estimation, we focus on a more challenging problem where CAD models are not available at inference time. Existing approaches mainly follow a complex multi-stage pipeline which first localizes and detects each object instance in the image and then regresses to either their 3D meshes or 6D poses. These approaches suffer from high-computational cost and low performance in complex multi-object scenarios, where occlusions can be present. Hence, we present a simple one-stage approach to predict both the 3D shape and estimate the 6D pose and size jointly in a bounding-box free manner. In particular, our method treats object instances as spatial centers where each center denotes the complete shape of an object along with its 6D pose and size. Through this per-pixel representation, our approach can reconstruct in real-time (40 FPS) multiple novel object instances and predict their 6D pose and sizes in a single-forward pass. Through extensive experiments, we demonstrate that our approach significantly outperforms all shape completion and categorical 6D pose and size estimation baselines on multi-object ShapeNet and NOCS datasets respectively with a 12.6% absolute improvement in mAP for 6D pose for novel real-world object instances.

PDF Abstract


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
6D Pose Estimation using RGBD CAMERA25 CenterSnap mAP 10, 10cm 87.9 # 1
mAP 10, 5cm 81.3 # 1
mAP 3DIou@25 93.2 # 1
mAP 3DIou@50 92.5 # 1
mAP 5, 5cm 66.2 # 1
6D Pose Estimation using RGBD REAL275 CenterSnap mAP 10, 10cm 70.9 # 2
mAP 10, 5cm 64.3 # 5
mAP 3DIou@25 83.5 # 6
mAP 3DIou@50 80.2 # 5
mAP 5, 5cm 29.1 # 7


No methods listed for this paper. Add relevant methods here