Thanks for sharing the MAS3K dataset. However, I find there are no source code. Besides, there are very different results between your paper and the following work[1]. The metrics are the same, but the results are very different. Could you explain this? Or, could you provide the segmentation mask predictions of other compared methods in your work? Then, one could fairly compare with.
[1] A Robust Object Segmentation Network for Under Water Scenes, ICASSP2022 (https://ieeexplore.ieee.org/abstract/document/9746176/)
Thanks for sharing the MAS3K dataset. However, I find there are no source code. Besides, there are very different results between your paper and the following work[1]. The metrics are the same, but the results are very different. Could you explain this? Or, could you provide the segmentation mask predictions of other compared methods in your work? Then, one could fairly compare with.
[1] A Robust Object Segmentation Network for Under Water Scenes, ICASSP2022 (https://ieeexplore.ieee.org/abstract/document/9746176/)