About the ProtoNet with Conv4 backbone result

[johnnyasd12]

Hi, I re-do the Conv4-ProtoNet 5-way-5-shot experiment, but the accuracy is about 4% more than the paper.

python ./train.py --dataset CUB --model Conv4 --method protonet --train_aug
python ./save_features.py --dataset CUB --model Conv4 --method protonet --train_aug
python ./test.py --dataset CUB --model Conv4 --method protonet --train_aug

I ran several experiments with the same setting, getting results with 74-75% accuracy. But the paper shows the result with 70.77%, Is there something wrong?

Thanks!

[rohitgajawada]

Yeah, even I get similar results

[bfshi]

me too

[bjlgcxc]

me too, even up to 75.72% +- 0.66%

[wyharveychen]

Hello, sorry about my reported the results for CUB dataset on the paper run fewer episodes and thus not fully converged. I have updated the fully converged results on the Arxiv as the follow link https://arxiv.org/abs/1904.04232.