Lane-Detection-In-Foggy-Weather

Project work submitted part of Deep Learning: Applications in Engineering at Clemson University - International Center for Automotive Research

Drivable Area Detection in Inclement Weather conditions.

Results:

• mIOU = 44.15%
• Architecture: Convolutional Neural Net
• Dataset: From BDD100K only inclement weather images - Total of 15630

Environment and Dependency

• Python 3.6
• Tensorflow 3.5
• OpenCV (Compatible with Python 3.6)
• CUDA 11.0.3 cuDNN 8.0 for Tesla V100 - 16GB
• HPC Cluster - Palmetto Cluster, Clemson University

Demo Video

Demo Video