hualin95 / Deeplab-v3plus

A higher performance pytorch implementation of DeepLab V3 Plus(DeepLab v3+)

hualin95/Deeplab-v3plus Issues

train issues
Updated 3 years ago2
fail to reproduce 79.19% IoU on val using train_aug
Updated 3 years ago2
GPU crashes while training, GPU is lost error!
Updated 3 years ago1
There are no train.py and libs.
Updated 4 years ago4
Can you provide a BaiduYun link of the pre-trained deeplab-v3 plus model ?
Closed 6 years ago3
pretrained weight for cityscape
Updated 5 years ago
I got a very low MIoU, MPA1 value when I started training. Only 0.03
Closed 6 years ago2
About hyper-params and frozen bn
Updated 5 years ago
Results on the PASCAL VOC dataset
Updated 5 years ago
Fail to produce the 79.19% result.
Closed 6 years ago3
Fail to reproduce the reported result.
Closed 6 years ago7
how about the cityscapes result ?
Updated 5 years ago7
How to test the result?
Updated 5 years ago
RuntimeError: cannot join current thread
Closed 5 years ago1
ASSP weights and bias did not update
Closed 6 years ago1