Results on PhotoTour about ral-net HOT 5 CLOSED

On Fri, Apr 5, 2019, 10:56 PM yanwu xu ***@***.***> wrote: The results are within the normal range (differences to our results are around 0.05). Since the algorithms are trained using SGD, the results cannot be the same as those reported in the paper. We used the model trained from Liberty data, and the details are in the paper. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AmwA-MhN_PmLYZCBrxPxgXwQu0ei2Bdkks5vd445gaJpZM4cPZc8> .

On Fri, Apr 5, 2019, 11:19 PM yanwu xu ***@***.***> wrote: RAL_loss_9_5000000_512_liberty.th.zip <https://github.com/xuyanwu/RAL-Net/files/3049146/RAL_loss_9_5000000_512_liberty.th.zip> — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AmwA-Nj7wIGOmZi6xq2cP3V2hatiR0ixks5vd5OygaJpZM4cPZc8> .