Qibin Hou
Qibin Hou
Thanks for your interests in this paper. I have tried spnet50 on Cityscapes and the mAP score is 79.5 on the validation set (w/o hard example mining). Please check your...
Hi, The local context aggregation is conducted in the value projection operation. The attention weights are generated by a linear but an unfold operation is operated on the value tensor....
> Hi thank you for your paper and congrats on SOTA. > > I have a question related to this, from the linear projection we generate an attention map for...
Thanks for your question. The difference is clear. The outlooker in VOLO is a new attention mechanism that targets at encoding fine-level token representations. We use a linear to generate...
Hi, thanks for sharing your log with us. Did you train any other models with your dataset and are they normal?
Have you ckeched that there is no problem with your val set?
Seems that there are some problems with the training data
Thanks for your comments. This is really important. As you suggested, I will release the training log soon.
I do use, but it depends on your hardware. Sometimes, apex amp works better in terms of training efficiency.
Two 1x1 convs are used to build the inter-channel relationship.