juhongm999
commented
2 years ago Did you train on 1 shot and used this model for getting 5-shot or did you train again using 5-shot ?
- Yes. We train our model in 1-shot setting and show that the trained model can be used for n-shot setting without loss of generality in Section 4.5 of our paper.
and the figure A4 in the paper is from resnet101 backbone?
- Yes. All the qualitative results in the paper (including supp.) are produced by model with resnet101 backbone.
moonsh
Hi, first, thank you for sharing your great work !.
I just wonder how you got the 5-shot results for tables 1,2 and 3 in the paper.
Did you train on 1 shot and used this model for getting 5-shot or did you train again using 5-shot ?
and the figure A4 in the paper is from resnet101 backbone?
Thank you.