Label-efficient Segmentation via Affinity Propagation
This addresses the problem of reducing labeling costs for segmentation tasks, though it appears incremental as it builds on existing pairwise affinity techniques.
The paper tackles weakly-supervised segmentation with sparse annotations by formulating affinity modeling as an affinity propagation process, achieving superior performance on three tasks including box-supervised instance segmentation and point/scribble-supervised semantic segmentation.
Weakly-supervised segmentation with label-efficient sparse annotations has attracted increasing research attention to reduce the cost of laborious pixel-wise labeling process, while the pairwise affinity modeling techniques play an essential role in this task. Most of the existing approaches focus on using the local appearance kernel to model the neighboring pairwise potentials. However, such a local operation fails to capture the long-range dependencies and ignores the topology of objects. In this work, we formulate the affinity modeling as an affinity propagation process, and propose a local and a global pairwise affinity terms to generate accurate soft pseudo labels. An efficient algorithm is also developed to reduce significantly the computational cost. The proposed approach can be conveniently plugged into existing segmentation networks. Experiments on three typical label-efficient segmentation tasks, i.e. box-supervised instance segmentation, point/scribble-supervised semantic segmentation and CLIP-guided semantic segmentation, demonstrate the superior performance of the proposed approach.