U2PL issues

Train custom data

3

Can you explain how to train custom data please ?

Holmes2002

error in running u2pl wth AEL

1

HI When running U2PL with AEL. I am getting this error: Traceback (most recent call last): File "../../train.py", line 967, in main() File "../../train.py", line 74, in main model =...

prantikhowlader

训练过程

4

您好，请问您能将VOC2012的1/8训练过程中生成的.txt文件给我看一下吗？

DeeplearningXiaobai

About ablation study on alpha_0

4

Hi there, I notice that you ablate alpha_0 in Tab.7 of paper, showed that 20% is the best setting for your work. Firstly, the top-20% low confidence for negative samples...

DISAPPEARED13

Using pre-trained weights from vision transformer

3

Hi there, thanks for the great tool! Would it be possible to use pre-trained weights from a vision transformer, instead of resnet? I had a quick go, but quickly ran...

elle-miller

How to split the datasets (cityscapes) into labeled and unlabeled?

2

Hi, nice work and I am going to reproduce the experimental results. I want to know how you split the dataset into the labeld and unlabeled parts? Is it random...

ywher

关于DPA与threshold的区别

您好！请问DPA和fixed threshold的训练曲线在训练前期会有差别吗？我目前正在尝试用DPA替代阈值，但是发现在训练前期DPA的mIOU曲线显著低于使用阈值下的曲线。请问这种情况是正常的吗？

Hugo-cell111

关于两种数据集训练的差异问题

5

您好，最近我在您的项目上进行修改并学习，目前在pascal voc上效果较好，但是在cityscapes数据集上效果很差，目前找了很久没找到原因，于是想问您一下，请问您在训练两种数据集上有什么设置上或者训练策略上需要注意的地方呢？目前我看到了一些config文件内的细小差别，可是怎么尝试效果都一般，如果能回复并给我一些建议的话，我将非常感谢！

syqisme