image-segmentation-keras icon indicating copy to clipboard operation
image-segmentation-keras copied to clipboard

Implementation of Segnet, FCN, UNet , PSPNet and other models in Keras.

Results 125 image-segmentation-keras issues
Sort by recently updated
recently updated
newest added
trafficstars

Hey, great work! I think these implementations are interesting. I am using this repository to work on my thesis, and I would like to know wheter it exists any official...

Hey guys, I have a dataset with only 3 classes ( background, class_1, class_2). The problem is, that the first 3 class colors look really similar to each other. Is...

Colab has upgraded to tensorflow 2.5 and now your examples are broken.

I am working on satellite image segmentation, I am getting the following error ![Screenshot from 2021-04-24 15-44-26](https://user-images.githubusercontent.com/67013108/115959307-4924a780-a514-11eb-85ba-35b167672137.png)

I have 2 error: 1.Error in loading augmentation, can't import imgaug.Please make sure it is installed. When I run command in cmd (Windows10): python -m keras_segmentation train \ --checkpoints_path="path_to_checkpoints" \...

Hi, I trained the model as shown in the tutorial and noticed that the default loss function is cross-entropy. Is there a way I can train the network with other...

Hello, this code seems to use only 512 training images/examples per epoch of training. As defined in train.py, steps_per_epoch = 512. I have 3352 training images in total. When I...

Hi, I am using two images as input. I am providing 2nd image using ` other_inputs_paths=[ "/path/to/other/directory" ]` But by doing so I am getting an error: `ValueError: Input 0...

I want train images with different size model = unet_mini(n_classes=2 , input_height=None, input_width=None ) ` 87 input_width = i_shape[2] ` ` 88 n_classes = o_shape[3]` `---> 89 o = (Reshape((output_height*output_width,...

The input and output format of UNet mini are (*, 360, 480, 3) and (1, 1, 172800, 50). We need to know how to use this output for the generation...