A segmentation model specialized in detecting and segmenting speech bubbles in comic and manga images.