PyTorch Image Segmentation Tutorial with U-NET: everything from scratch baby

How did you do the masking in the dataset? How did you create the dataset, where can I learn the detailed explanation?

Carvana kaggle dataset does not seem to have val_images and val_mask

thank you for this video! after watching a handful of times, I've managed to get it predicting on my own custom dataset, thanks entirely to your instruction.
curious though - any advice on where to start getting a successful model to make a prediction on a single image, and call it by a script?

Thank you add. where is this source code? thanks.

Dear professor,

I am very interested in your program, and I have two questions now,

(1) How to use code to map between irregular images, complete training through the unet model, and then conduct testing?

Is the mask used for preprocessing data? Is there any special software available for preprocessing?

learnt soo much from this thank you! love the proper structure instead of line by line commands in colab or sth

Thanks! Great work. Useful practical information

how can i test implemented model when test dataset is given??? anyone help ...

thanku so much the explanations made it very clear 🙌💯

i followed your tutorial step by step and used the same dataset and it did an amazing job. The first dataset (CARVANA) I used worked fine, but once I changed it, the results went downhill. I tried it on CASIAv2, but my dice score is always 0.0 and my predicted masks are just black... i don't know how to fix this, if anyone has any ideas, i beg you, do let me know!

Thank you so much my guy. I hope one day I can also do this with my own knowledge and understanding

@AladdinPersson
What kind of PyCharm theme do you use? Looks awesome!

Anyone else getting a consistent accuracy every epoch, a dice score of 0 and empty black images for the prediction? Don't know what I could be missing

so, what changes do i need to make if I want to perform a multi class segmentation here can you help me?

I have a problem, import torchvision.transforms.functional gives module error and says it is not a library

Hey there, thanks for making this video.
I tried running this code but it's popping a type error while running the model part of the it.
TypeError: unsupported operand type(s) for %: 'list' and 'int'
Tried looking up for possible solutions but didn't help.
Would greatly appreciate if you could weigh in on this.
Thank you!

Great video, man!

Very good explanation using pytorch and Unet, I was able to use that in 1024x1024 images but with 416x416 your DICE formula always shows 0.0, even if I have 99% accuracy, I don't know why...please one suggestion, thanks

It's unbelievable that I got the "nan" loss in me laptop(NVIDIA 1650), but a pretty good result in the remote server(3080ti) with the same code.😅

Hi everybody, why my acc data is 78.20....? And saved_images/pred_x.png are all black picture?

what a great tutorial

Unfortunately, it doesn't work with .tif files :(

Thanks a ton!!!!! Learnt a hell lot of new things from this video other than image segmentation.
Your lectures are pure gem!!!!

hi, may i ask you something
do i have to change the IMAGE_HEIGHT and IMAGE_WIDTH if i use my own dataset?

What do I have to modify here if I've got 8 output classes rather than binary segmentation?

Hey @Aladdin Persson here for binary classification you applied sigmoid to the outputs of the model and then just separated into two by threshold of 0.5, can you suggest anything similar for multiclass classification? can softmax be used there? if yes, how can i separated then further?

This is a very well done tutorial

Validation set is not available on Kaggle

Binary cross entropy? Don't you still need to use cross entropy as the loss function because each pixel is a "class?"

Thanks for this Aladdin. I was able to train using my own data. Do you have an idea how I can deploy U-net model to my web app? Can't seem to find any resource on it. cheers

thank you so much for this content

Hi. Thank you for your video. It helped me a lot

Thanks for creating this education video. Every concept is very clearly explained.

How do you create a mask image file and can you create it yourself?

Can we only use this if we have the masks in the train dataset ?

Hello, I implemented it but I don't know why my loss is going negative after each epoch starting from -46 to -150. Can you help what might be the cause . its binary segmentation.

Thanks for the video. Why you used scaler for backward ? I did not totally understand that.

You go too fast man, had to play it on .75x

very good video

Thanks for the tutorial.

Hmm, that trick you added to avoid the requirement of having input perfectly dividable by 16 might lead to big issues depending on the type of imagery that is being processed by the network. Imagine satellite imagery with a GSD (ground sampling distance) of 100m. A single pixel is literally 100x100m and skipping one leads to skipping multiple houses. :D Just saying this in case people come across your tutorial and just blindly copy paste the code.

NOTE: Kaggle requires phone number for verifying your account. For those of you (like me), who do not want to hand out such private information, find another set. In the end U-Net is used in many fields with different types of images (e.g. medical ones) and the chances are you will not be doing segmentation on cars. :D

Great video man. You are working with RGB images (3 bands or channels). Do you think is possible use this architecture for images with more than 3 channels or bands. I'm thinking in hyperspectral cameras, for example.

Thanks for the video

Thank you so much for this informative and detailed tutorial.

Hey aladdin my dice score is going over 1 Idk why that is, please help!!!

Many thanks of writing this specifically with PyTorch from scratch, I love your videos doing from scratch, you are awesome

thank you so much,I learnt a lot from this vedio. You are awesome!!!