# Changelog

All notable changes to this project will be documented in this file.

The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

## [Unreleased]

## [0.1.1] - 2026-01-16

### Added
- `useVideoScanner` hook for real-time video scanning with temporal smoothing
- `useReceiptScanner` hook for receipt OCR with MLKit text recognition
- TFLite frame processor for on-device real-time detection
- YOLO output processing utilities (`yoloProcessor.ts`)
- Image preprocessing utilities with Skia support (`imagePreprocessor.ts`)
- Detection overlay component
- COCO class labels (80 classes) support
- SousChef custom model (233 food classes) support
- Configurable logging utility

### Changed
- Updated `useProductDetector` with image preprocessing via Skia
- Improved model loading with Android asset fallback
- Enhanced server detection with detailed logging

### Fixed
- Global model state management with singleton pattern
- Event listener cleanup in ServerClient
- Unused variable warnings

## [0.1.0] - 2026-01-15

### Added
- Initial release
- `MLVisionProvider` context for configuration and caching
- `useMultiBarcodeScanner` hook for multi-barcode detection
- `useProductDetector` hook for visual product recognition
- `CacheManager` with MMKV-based caching
- `ServerClient` for fallback inference server communication
- TypeScript type definitions for all features
- Docker training environment for YOLOv8 models
- Documentation and architecture guides

### Technical Details
- React Native 0.73+ support
- TFLite INT8 quantized models (~3-6MB)
- On-device MLKit barcode scanning
- GPU delegate support (CoreML on iOS)

[Unreleased]: https://github.com/SousChefLabs/ml-vision/compare/v0.1.1...HEAD
[0.1.1]: https://github.com/SousChefLabs/ml-vision/compare/v0.1.0...v0.1.1
[0.1.0]: https://github.com/SousChefLabs/ml-vision/releases/tag/v0.1.0