Not achieving the accuracy stated in the paper

hustvl / WeakTr

WeakTr: Exploring Plain Vision Transformer for Weakly-supervised Semantic Segmentation

MIT License

120 stars 2 forks source link

Not achieving the accuracy stated in the paper #19

Closed wsgsgg closed 9 months ago

wsgsgg commented 11 months ago

Hello author, why did I reproduce your model on the Pascal dataset with a final miou of only 74.01%

wsgsgg commented 11 months ago

I don't know what the reason is. I hope to receive your help

Unrealluver commented 11 months ago

Hi! Could you please provide more details of your experiments so we can help you analyze the possible problem? And you could try our checkpoints and pseudo masks here.

Unrealluver commented 9 months ago

I will close this issue. If you have more questions, you are welcome to raise issues :)