Hao Wang (王浩) comments

Results 6 comments of


                                            Hao Wang (王浩)

can't reproduce accurate MAP

(1) I got 0.058 map reproduced, too. (2) I think the word xxx should be seperated, for example, car_(sedan)

can't reproduce accurate MAP

@payne53 Thank you so much for pointing it. Could you reproduce the performance on QuickDraw dataset?

The performance is lower than the results in your paper

How about the random seed? Did you use fixed seed or randomly initialize, such as np.random function in data_loader?

The performance is lower than the results in your paper

Thanks for your reply. Maybe I use the different versions of PyTorch or CUDA.

The performance is lower than the results in your paper

(1) Would you share the random seed for training? (2) How about the environment you used? I tried PyTorch 0.4.1 and Python 3.5, nothing changed. The data_loader maybe the key...

The performance is lower than the results in your paper

Anyone who can reproduce the results of AwA1 and other datasets? I am very appreciate for your help!