ibaiGorordo/TfLite-Ultra-Fast-Lane-Detection-Inference ? reverse-engineered prompt

Build me a small Python project that can detect road lane lines using the Ultra Fast Lane Detection TensorFlow Lite model.

I want it to work on a single image, a webcam feed, and a video file, with each mode in its own simple script. The app should load the TFLite model from a models folder, run inference, and draw the detected lanes on top of the original frame so I can clearly see the results. If there’s a sample image included, use that for the image demo, and make the webcam and video versions easy to run from the command line.

Please keep the setup straightforward, with a requirements file and clear instructions for installing what’s needed. If any extra packages are needed for YouTube video input or TensorFlow Lite runtime, include that too. Make the code easy to understand and practical for running on a laptop or a Raspberry Pi. If you need to check current docs or model details online, go ahead and do that.