segdec-net-jim2019
segdec-net-jim2019 copied to clipboard
skokec
Surface Defect Detection with Segmentation-Decision Network on KolektorSDD
It's great project, thanks for your share. in some cases, I want to get precise contours of the predicted defects, how to adjust the config parameters?
Could you specify what kind of license do you permit to use the code and the dataset?
Hi, I just tried to use the rotation=True option in the code however it seems not working: File "../segdec-net-jim2019/segdec_model.py", line 258, in get_loss l1 = tf.losses.sigmoid_cross_entropy(logits=predictions,multi_class_labels=masks) InvalidArgumentError (see above for...
Hello, Thanks for the great work. The training/evaluation code works perfectly with KolektorSDD (with dilate=5 for mask) dataset. But, when I run the training on my dataset, tensorflow ends up...
Hello, I was able to run this code ok with CPU, but got a problem running in parallel. The error message is as the following: Attempting to fetch value instead...
Hello, I read your paper Segmentation-Based Deep-Learning Approach for Surface-Defect Detection and amazed by your results. However, I couldn't understand very well what property of your network allows a training...
Hi, Thanks for the great work, but I have question about the decision network and how it help with segmentation. According to your paper, segmentation network is trained separately. When...
Amazing algorithm. I am using this algorithm on my own dataset, there is around 600 data images and a third of them have defects. But I keep experiencing expolding gradients...
Hello, really great work! Has this code been tested with tensorflow 1.14? Sinan
Hi, great work! I am trying to serializing the model into pb model format, but I can't figure out the output nodes' names of the model. Do you know what...
