Supplementary Material for Self-supervised Amodal Video Object Segmentation