Giters

deepplants / ViT-PCM

Official implementation of "Max Pooling with Vision Transformers reconciles class and shape in weakly supervised semantic segmentation"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Questions about the retraining network

Unrealluver opened this issue · comments

commented

Greetings!

Thanks for your released code for ViT-PCM. It's excellent work that shows the ability of ViT for WSSS. When I read your paper, I found no more details about the retrain method. Could you please tell me which pretrained weight you used in the experiments? Imagenet1k pretrained weight or COCO pretrained？

Looking for your reply!

commented

Hello @Unrealluver !
We used Imagenet-1K pertained weights in all experiments.
Thanks for your feedback,
Bye!

commented

@rossettisimone Thanks for your reply! Could you please share the retraining code?

commented

Hello @Unrealluver,
Actually, we have no custom retraining code, we used DeeplabV2+dCRF from kazuto1011 and weights initialisation from johnnylu305 and did some hyperparameters tuning.
Bye!