search

aurora95 / Keras-FCN

Keras-tensorflow implementation of Fully Convolutional Networks for Semantic Segmentation（Unfinished）
MIT License
650 stars 268 forks source link

GOT VERY LOW mIOU and Pixel ACC #61

Open XChuanLee opened 6 years ago

XChuanLee commented 6 years ago

I run train.py toward my own data_path _train_file_path = os.path.expanduser('/home/wang/Li_XC/KSFRCNN/data/VOCdevkit/VOC2012/ImageSets/Segmentation/train.txt') #Data/VOClarge/VOC2012/ImageSets/Segmentation val_file_path = os.path.expanduser('/home/wang/Li_XC/KSFRCNN/data/VOCdevkit/VOC2012/ImageSets/Segmentation/val.txt') data_dir = os.path.expanduser('/home/wang/Li_XC/KSFRCNN/data/VOCdevkit/VOC2012/JPEGImages') label_dir = os.path.expanduser('/home/wang/LiXC/KSFRCNN/data/VOCdevkit/VOC2012/SegmentationClass') But I got a very very low IOU matrix when I run evaluate.py. meanIOU: 0.014367 pixel acc: 0.124107 Could somebody helps?

daodaoawaker commented 5 years ago

@XChuanLee Are you successful in training? Are you meanIOU raised

daodaoawaker commented 5 years ago

@XChuanLee Hi, have your meanIOU and acc been improved later ?

gdut2hj commented 4 years ago

@daodao1991 I encountered the same problem, did you solve it?