3D Objectness Estimation via Bottom-up Regret Grouping
This addresses the problem of discovering semantic objects in 3D scenes for 3D understanding, representing an incremental improvement over prior bottom-up methods.
The paper tackles 3D objectness estimation by proposing a bottom-up method with a grouping predictor and regret mechanism to reduce errors from incorrect groupings, achieving state-of-the-art performance with fewer proposals on datasets like GMU-kitchen and CTD.
3D objectness estimation, namely discovering semantic objects from 3D scene, is a challenging and significant task in 3D understanding. In this paper, we propose a 3D objectness method working in a bottom-up manner. Beginning with over-segmented 3D segments, we iteratively group them into object proposals by learning an ingenious grouping predictor to determine whether two 3D segments can be grouped or not. To enhance robustness, a novel regret mechanism is presented to withdraw incorrect grouping operations. Hence the irreparable consequences brought by mistaken grouping in prior bottom-up works can be greatly reduced. Our experiments show that our method outperforms state-of-the-art 3D objectness methods with a small number of proposals in two difficult datasets, GMU-kitchen and CTD. Further ablation study also demonstrates the effectiveness of our grouping predictor and regret mechanism.