Can you share the model file that your Single-task Model trained on the verse2020 dataset, I can't achieve the effect in the paper?

[XFivezzz]

[yeerwen]

Did you add a deep supervision strategy to these single-task models and implement them based on the nnUNet framework?

[XFivezzz]

Thanks for your reply，Yes, I did，Have you measured the IoU, HD, Percision, and Recall metrics of these single-task models?I asked for the model trained in the experimental environment of the paper in order to better validate these indicators.

[yeerwen]

No, we did not measure other metrics. For the competitive approach, our modifications are (1) training the model based on the nnUNet framework; and (2) adding a deep supervision strategy to each decoder output. We report the key hyperparameters, such as learning rate and optimizer, in the appendix material.