AsyInst: Asymmetric Affinity with DepthGrad and Color for Box-Supervised Instance Segmentation
This work addresses instance segmentation for computer vision applications, but it is incremental as it builds on existing box-supervised methods.
The paper tackled limitations in weakly supervised instance segmentation using bounding boxes, specifically addressing issues with pairwise affinity loss, and proposed an asymmetric affinity loss that improved mask AP by 3.5% over the baseline on the Cityscapes dataset.
The weakly supervised instance segmentation is a challenging task. The existing methods typically use bounding boxes as supervision and optimize the network with a regularization loss term such as pairwise color affinity loss for instance segmentation. Through systematic analysis, we found that the commonly used pairwise affinity loss has two limitations: (1) it works with color affinity but leads to inferior performance with other modalities such as depth gradient, (2)the original affinity loss does not prevent trivial predictions as intended but actually accelerates this process due to the affinity loss term being symmetric. To overcome these two limitations, in this paper, we propose a novel asymmetric affinity loss which provides the penalty against the trivial prediction and generalizes well with affinity loss from different modalities. With the proposed asymmetric affinity loss, our method outperforms the state-of-the-art methods on the Cityscapes dataset and outperforms our baseline method by 3.5% in mask AP.