huijuan88
commented
5 years ago The learning rate at these layers can also be open. But the training time will increase. It is a trade-off.
On Jan 29, 2019, at 09:49, rambleramble notifications@github.com<mailto:notifications@github.com> wrote:
It seems that the 1st-5th group won't be changing weights during the training since there is a pre-trained model in your descriptionhttps://github.com/VisionLearningGroup/R-C3D/blob/master/experiments/thumos14/train.prototxt
"The C3D model weight pretrained on Sports1M and finetuned on ActivityNet dataset"
Is this a correct understanding? Can you provide a more comprehensive expalantion about "end-to-end", in terms of training via back-propagation? For instance, maybe we can use your paper's Fighure 2 as a reference. Thanks a lot!
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/VisionLearningGroup/R-C3D/issues/52, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AFOa_6ulddnCweImNq14YMqW0bIMTJ7Gks5vIKW1gaJpZM4aYzWA.
It seems that the 1st-5th group won't be changing weights during the training since there is a pre-trained model in your description below: "The C3D model weight pretrained on Sports1M and finetuned on ActivityNet dataset"
Is this a correct understanding? Can you provide a more comprehensive expalantion about "end-to-end", in terms of training via back-propagation? For instance, maybe we can use your paper's Fighure 2 as a reference. Thanks a lot!