Lavideol/Vehicle-Detection-Nepal

Debug: Add detailed step-by-step logging for background processing

Simplify: Remove problematic debug code, fix file handling

Fix: Add detailed upload debugging and handle model version incompatibility gracefully

Fix: Better error handling, lower detection threshold (0.3), improved logging

Debug: Add detailed logging for classifier loading

Simplify: Remove Git LFS from Dockerfile (using direct file now)

Add classifier model directly (remove LFS)

Fix: Add Git LFS support and improve classifier loading with debugging

Delete models/.gitkeep

Delete vehicle_classifier_final.pt

Upload vehicle_classifier_final.pt

Create models folder structure

Update: Sync app.py with improved error handling

Improve: Better model detection - use existing classifier.pt if available

Upload vehicle_classifier_final.pt

Optimize: Pre-download YOLO models during Docker build

Improve: Better error handling and user feedback for file uploads

Fix: Pin NumPy to 1.x to fix OpenCV compatibility

Fix: Use full Python 3.10 image with complete system dependencies for OpenCV

Fix: Update Dockerfile with available OpenCV dependencies

Fix: Add missing OpenCV graphics libraries to Dockerfile

Add Hugging Face Spaces configuration README

Vehicle Detection Nepal - Initial Deployment