Weighted Mask R-CNN for Improving Adjacent Boundary Segmentation

In the recent era of AI, instance segmentation has significantly advanced boundary and object detection especially in diverse fields (e.g., biological environmental research). Despite its progress, edge amid adjacent objects organism cells) still remains intractable. This is because homogeneous heterogeneous are prone to being mingled a single image. To cope with this challenge, we propose weighted Mask R-CNN designed effectively separate overlapped virtue extra weights boundaries. For numerical study, range experiments performed applications simulated data real Microcystis, one most common algae genera cell membrane images). It noticeable that outperforms standard R-CNN, given analytic show on average 92.5% precision 96.4% recall 94.5% 98.6% data. Consequently, found majority sample boundaries precisely segmented midst mixtures.