Rethinking Task and Metrics of Instance Segmentation on 3D Point Clouds
This work addresses a critical issue in 3D instance segmentation for autonomous cars and robots by improving evaluation accuracy and scalability, though it is incremental as it builds on existing methods with novel metrics and efficiency gains.
The paper tackles the problem of instance segmentation on 3D point clouds by addressing limitations in existing methods that split inputs into small regions due to high space complexity, which skews evaluation metrics like mAP. It proposes a new method with O(Np) space complexity to handle large regions and introduces novel metrics independent of categories or input size, achieving state-of-the-art performance.
Instance segmentation on 3D point clouds is one of the most extensively researched areas toward the realization of autonomous cars and robots. Certain existing studies have split input point clouds into small regions such as 1m x 1m; one reason for this is that models in the studies cannot consume a large number of points because of the large space complexity. However, because such small regions occasionally include a very small number of instances belonging to the same class, an evaluation using existing metrics such as mAP is largely affected by the category recognition performance. To address these problems, we propose a new method with space complexity O(Np) such that large regions can be consumed, as well as novel metrics for tasks that are independent of the categories or size of the inputs. Our method learns a mapping from input point clouds to an embedding space, where the embeddings form clusters for each instance and distinguish instances using these clusters during testing. Our method achieves state-of-the-art performance using both existing and the proposed metrics. Moreover, we show that our new metric can evaluate the performance of a task without being affected by any other condition.