Supplementary Material for SegRefiner: Towards Model-Agnostic Segmentation Refinement with Discrete Diffusion Process Anonymous Author(s) Affiliation Address email 1 Implementation Details 1

The overall workflow of the training and inference process are provided in Alg. 1 and Alg. 2. Model Architecture Following [ 9 ], we use a U-Net with 4-channel input and 1-channel output. Both input and output resolution is set to 256 256 . Training Settings All experiments are conducted on 8 NVIDIA RTX3090 GPUs with Pytorch. After a complete reverse diffusion process, the output is resized to the original size. We apply Non-Maximum Suppression (NMS, with 0.3 as threshold) to these patches to remove Our SegRefiner can robustly correct prediction errors both outside and inside the coarse mask.