search

ibrahimgh25 / EL-GAN-Implementation

A pytorch implementation for the network in "EL-GAN: Embedding Loss Driven Generative Adversarial Networks for Lane Detection" by Ghafoorian et al.
MIT License
3 stars 0 forks source link

Can you share the test code? #3

Open zxk72 opened 4 months ago

zxk72 commented 4 months ago

Thanks for your excellent work! I read your other answer, but I am not very good at it. I would like to ask if you can send me the code of only the test part? Looking forward to your reply, thank you very much!

ibrahimgh25 commented 3 months ago

Hello,

Unfortunately I don't have the "test" code. I tried running the project, a lot has changed since the code was written, I can't even find pytorch 1.8.0 on pypi. In any case, if you were able to somehow download the project, the data, and run the training script you're already there. the testing would be predicting with the trained model (generator) and comparing your results with the ground truth, you could use the loss function (pixel-wise cross-entropy), or this much better evaluation script from the challenge: evaluate.py.

I might in the coming days/weeks work on updating the project to work with python 3.10 (I was working on 3.6 at the time I think), and pytorch 2.1, but for that I will have to rewrite a lot of the code, since it's messy. I might try training the model again, if I do I will let you know.

zxk72 commented 3 months ago

@ibrahimgh25 Thank you very much for your reply. I did encounter some troubles running the existing code, and I am trying to figure it out. In the meantime, I look forward to your updates, and I am very grateful for your updates on the code, which are like treasures to me. Thank you!