Hello, I would like to ask whether the training results on the PolypSeg dataset reported in your paper were obtained from multiple experiments to find the best results for different models, or if they represent the results of a single best model on different test sets of PolypSeg. Due to variations in random seeds, we are unable to fully reproduce the results reported in the paper. We have found that even the best-performing model shows significant performance differences across different test sets within the PolypSeg dataset. We are unable to obtain the best results on different test sets of PolypSeg from a model trained in a single run.I hope to receive your guidance and further clarification. Thank you!
Hello, I would like to ask whether the training results on the PolypSeg dataset reported in your paper were obtained from multiple experiments to find the best results for different models, or if they represent the results of a single best model on different test sets of PolypSeg. Due to variations in random seeds, we are unable to fully reproduce the results reported in the paper. We have found that even the best-performing model shows significant performance differences across different test sets within the PolypSeg dataset. We are unable to obtain the best results on different test sets of PolypSeg from a model trained in a single run.I hope to receive your guidance and further clarification. Thank you!