Releases: nischi/MMM-Face-Reco-DNN
Releases · nischi/MMM-Face-Reco-DNN
Suspend Recognition
New Features
- Conditional triggering of face recognition based on Magic Mirror notifications.
- New configuration option external_trigger_notification for enhanced control over face recognition activation.
Improvements
- Enhanced efficiency in face recognition processes, reducing CPU utilization and power consumption.
- Updated installation instructions for clarity, including SWAP space requirements and compatibility notes.
Bug Fixes
- Improved handling of shutdown signals for more reliable termination of processes.
Fixing dependencies
- Change dependencies from
opencv-python-headless
toopencv-python
- Add an image capture script. Thanks to @camillemarius
Remove pipenv
- After a lot of install issues. We removed
pipenv
again and changed the install-instructions- For Bookworm we added an extra install step for an virtual environment
Use new libraries to support new cameras
Use the new picamera2 module to get the images from the camera
- Using pipenv to load all packages
- Automatic install the dependencies
- Support new OS
- Support new Cameras
- Support 64bit
- Adding new support for contrast and brightness
- Change defualt resolution
- Add support to show picture in mirror
Image brightness & contrast adjustment
Bugfixes & new Features
- Better instructions
- Makefile for Docker
- ESLint and Prettier integrated
- Add new source parameter
- Add new rotateCamera parameter
- Add new tolerance parameter
- Better data handling
- Display face distance in tools
- Bugfix tolerance format
- Add russian
Bugfixes
Some new stuff :)
- Fix issue where process stayed open after shutdown from MagicMirror
- Fix issue json logging
- Update Readme for better install guide
- Added italian translations
- Add file path on encoding
Extend Dataset
New Feature implemented which extend your dataset with images each time a known or unkown people will recognize. Retrain need to be done manually.
Add Bulgarian
Add Bulgarian to the alert messages. Add Notification for Alert.