Training failed with yolo_segmentation

[abdou31]

Hello , I tried to train my own dataset using this project but it failed , For the first time , i didn't get any errors when runing this command :

./darknet segmenter train data/obj.data cfg/segment.cfg segment.backup this is the output for the first time ( but training doesn't work ):

33 logistic x entropy 173056 Loading weights from segment.backup...Done! the net n is 34 Learning Rate: 1e-07, Momentum: 0.9, Decay: 0.0005 2357 Saving weights to backup/segment.weights

When i tried again , i get this error :

   33 logistic x entropy                             173056
   Loading weights from segment.backup...Done!
   the net n is 34
   Learning Rate: 1e-07, Momentum: 0.9, Decay: 0.0005
   2357
   Saving weights to backup/segment.weights
   Cannot load image "/home/starinfo/yolo_segmentation/mask/Img_810_img.txt"
   STB Reason: can't fopen

How can i solve those issues ? Thanks

[tuuzhang]

@abdou31 According to the error message, you should load the image(ori and lable),while in fact you load a txt file instead! So you should check your obj.data file and train.txt(or train.list), the train.txt list the training image path and name!

[abdou31]

That is not that my error, You see my error is same that is in this link :

https://github.com/zhengshoujian/darknet-yolo-segmentation/issues/8

[ArtyZe]

Just as @tuuzhang said, the label is also a picture. And the problem occurs because you have to modify the function of load_image_gray according to your label picture name :)