search

hlamba28 / UNET-TGS

Applying UNET Model on TGS Salt Identification Challenge hosted on Kaggle
https://towardsdatascience.com/understanding-semantic-segmentation-with-unet-6be4f42d4b47
MIT License
142 stars 86 forks source link

normalization code for image data in mask directory does not work #8

Open superbunny38 opened 2 years ago

superbunny38 commented 2 years ago

Hello, thank you so much for your nice and detailed explanations about U-Net and Image segmentation. I had no idea what those are, and I was able to learn a lot from your thorough explanations and well-explained codes. However, when normalizing image data by dividing all images by 255.0, it does not work for images in mask directory. So when I fit the model, it gives me negative values for binary cross_entroy error :) So, I added the code myself to manually normalize image data in mask directory. Thanks for reading!

Pottah007 commented 2 years ago

Hello there, I'm facing the same issue. Could you please explain or provide the code which you used to properly normalize the data?

superbunny38 commented 2 years ago

Hello,

Hello there, I'm facing the same issue. Could you please explain or provide the code which you used to properly normalize the data?

Hello, @Pottah007 I will leave you the link below that I re-produced this work, and in which I also successfully handled the normalization issue. :) Here is the link: https://github.com/superbunny38/DeepLearning/blob/main/papers/U_Net.ipynb