Training of image based few shot methods

[nankepan]

Hi, how to get results(Table 3) of image based few shot methods(PMMs, PFENet and PPNet) in paper? Should I use their pretrained models to test on Youtube-VIS dataset, or train their models on Youtube-VIS dataset firstly? If training on video dataset is necessary, is it similar with DAN? could you give me more detail about it? Thank you!

[scutpaul]

Thanks for your interest. We train those models on Youtube-VIS. You can use our training code, the training process is similar to training our DAN model. The training config of those models can be referred to their public settings.