YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/model-cards#model-card-metadata)

Lane Detection with U-Net on TuSimple Dataset

This repository contains a trained U-Net model for lane detection using the TuSimple dataset.

Model Overview

  • Architecture: U-Net
  • Input shape: 224x224 RGB
  • Loss Function: Dice Loss
  • Metrics: Accuracy, Dice Coefficient, Precision, Recall

Contents

  • best_model.keras: Trained model weights
  • training_history.csv: Loss and accuracy logs
  • training_history.svg: Accuracy and loss curves
  • driving_car_lanes.mp4: Inference demo on driving footage
  • prediction_plots/: Sample predictions from the test set
Downloads last month
8
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support