LiteALPR: A lightweight open-source library for fast and accurate license plate recognition
Abstract
What's Changed in v0.1.7 🚀 Jetson Nano & Performance Benchmarking Jetson Nano Guide & Stress Testing: Introduced a comprehensive Jetson Nano deployment guide (docs/jetson-nano.md) along with a dedicated continuous stress-testing script (run_stress_test_jetson_nano.py) to validate thermal performance and memory footprint on Edge devices. Standardized Latency Benchmarks: Added the benchmark_latency.py script for precise end-to-end FPS measurement and profiling on desktop environments (CPU/GPU). Environment Reference: Added reference-environment.md to cleanly document hardware/software combinations for reproducible testing. 📓 Demo Notebook Enhancements Notebook Updates: Refined and improved all example notebooks (demo.ipynb, demo_cpu.ipynb, demo_gpu.ipynb) to ensure smoother user onboarding and clearer inference visualizations. 📚 General Documentation Refinements Extensive Docs Update: Heavily updated README.md, benchmarks.md, getting-started.md, and training.md with clearer instructions, more detailed benchmark tables, and better navigation flows in MkDocs. 📦 Package Maintenance Code Quality: Enforced strict ruff linting and formatting rules across the repository. Dependency Management: Added requirements.lock to ensure strictly reproducible environments. Bumped package version to 0.1.7 across litealpr, pyproject.toml, setup.py, CITATION.cff, and GitHub Action workflows. Full Changelog: https://github.com/vn-anhnth/LiteALPR/compare/v0.1.6...v0.1.7
