GitHub - pranav1601/lane-detection-with-attention: Built a CNN model using UNet architecture with Residual Blocks and Attention to detect lanes on US highways

Lane Detection using attention.

This repository contains the notebook lane_detector.ipynb that analyses two models - the UNet and a custom UNet. The Custom Unet utilizes residual blocks and attention to focus on more important features of the image.
I have used the TUSimple dataset. You can get it here.
you can access the Unet model and the custom UNet model here. Feel free to try them out
you can access the output videos here
You can view a detailed explanation of my algorithm on medium.

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
README.md		README.md
lane_detector.ipynb		lane_detector.ipynb
output1.png		output1.png
output_collated_video.gif		output_collated_video.gif