Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers
This addresses instance segmentation problems in computer vision, particularly for scenarios with heavy object occlusion, representing a novel method for a known bottleneck.
The paper tackles the challenge of segmenting highly-overlapping objects by modeling image formation as two overlapping layers, which decouples occluding and occluded instance boundaries. It achieves large and consistent performance gains, especially in heavy occlusion cases, as validated on COCO and KINS datasets.
Segmenting highly-overlapping objects is challenging, because typically no distinction is made between real object contours and occlusion boundaries. Unlike previous two-stage instance segmentation methods, we model image formation as composition of two overlapping layers, and propose Bilayer Convolutional Network (BCNet), where the top GCN layer detects the occluding objects (occluder) and the bottom GCN layer infers partially occluded instance (occludee). The explicit modeling of occlusion relationship with bilayer structure naturally decouples the boundaries of both the occluding and occluded instances, and considers the interaction between them during mask regression. We validate the efficacy of bilayer decoupling on both one-stage and two-stage object detectors with different backbones and network layer choices. Despite its simplicity, extensive experiments on COCO and KINS show that our occlusion-aware BCNet achieves large and consistent performance gain especially for heavy occlusion cases. Code is available at https://github.com/lkeab/BCNet.