martinkersner / train-DeepLab

Train DeepLab for Semantic Image Segmentation
MIT License
172 stars 76 forks source link

Different mean_iou on VOC12 #37

Open dingfg opened 6 years ago

dingfg commented 6 years ago

I tested resnet-101 successfully, but it's not the same as the author's miou. The result before densecrf on the validation set was 0.7646 and the paper was 0.7635. Strangely, the miou I ran out of VGG-16 was consistent with the paper. Does anyone have the same problem?

zhwis commented 6 years ago

What did you use for evaluate? Is it evaluate_deeplab.py. I use VOCdevkit to evaluate result,but their value of iou is not the same. Thanks