asdfqwer2015

Results 12 comments of asdfqwer2015

Hello, I used your resnet50 with ibna added, and the performance on the mixed data set is significantly reduced. What is the reason for this? The data set is multiple person reID data sets

Hi, I met a similar issue, have you solved this issue? @wfchenwf I used a mobilenet backbone in detection task, and added IN(with affine) after first conv and added IN...

commission was not considered

Same issue.

WT in the paper leaks info

Hope to see the amazing results. Please keep refining :)

overfitting very fast?

ok, I'll try BN istead of LN now. But I only have a single gpu to train the model, so the batchsize can be max at 4 for my scenario,...

[Question] Stacked-LSTM worse than Single LSTM (PPO2)

Does it works? @RonaldJEN

Obtained test accuracy lower than reported for miniImageNet

I got same issue too, even lower. But I tested Conv64 in my test, which got novel: 55.47% and base: 67.98% for posted novel: 55.86% and base: 68.43% by the...

Could you please give me some advices for improving acc_both?

Thanks for your reply and sorry for my late response, my network connection is not very well. I modified feature net and dataloader, and slightly modified FewShot due to newaxis...

Could you please give me some advices for improving acc_both?

En, 5 classes of gestures are chosen as novel classes. In stage2, AccuracyBase and AccuracyNovel have not increased, and I think AccuracyNovel has already very high. But AccuracyBoth dropped from...

Could you please give me some advices for improving acc_both?

Thanks for your help. Waiting for your advice. :)

no_action label in test mode

Thanks for your quick reply. I might haven't described the issue clearly. En, for buffer mechanism, I've noticed the buffer mechanism in your `ActionRecognition.py` script before. And I just modified...