Releases: fuzun/iot-facerecognition
Releases · fuzun/iot-facerecognition
Major Update
Server ⇾ v2.0.0
- Built with Qt 5.15.2 & MSVC.
- Qt runtimes are included.
- CUDA is NOT supported, you must build the server yourself if you want to make dlib use CUDA.
- A CPU with AVX, SSE2, SSE4 support is required to run.
- Requires Visual C++ Redistributables! They are not included. (MSVCP140, VCRUNTIME140, as well as win-crt)
- Requires libjpeg-turbo to be installed! (turbojpeg.dll)
- Built with Intel MKL support. The runtime is included.
- License is changed to AGPLv3 from GPLv3.
Client/qt ⇾ v1.0.0
- Initial release.
- A brand-new client made with C++ & Qt Framework that is intended to work on low power embedded devices such as Raspberry Pi 0 W.
Windows - x64
- Built with Qt 5.15.2 & MSVC.
- Qt Runtimes are included.
- Requires libjpeg-turbo to be installed! (turbojpeg.dll)
- Requires Visual C++ Redistributables! They are not included. (MSVCP140, VCRUNTIME140, as well as win-crt)
Raspberry Pi OS (Lite) - ARMv6
- The application requires a touch display connected through HDMI. No desktop environment / window system is required since EGLFS is used.
- The application requires a Raspberry Pi Camera module attached. The OpenMAX backend must be able to recognize the camera.
- Because of backwards compatibility, it will also run on machines with ARMv7 or later architecture (Raspberry Pi 3, 4, etc.).
- Qt Runtimes 5.15.2 are NOT included. Please install the following libraries (you can use the package manager) (note that they should ideally be 5.15.2 / a symbolic link to 5.15.2, I don't know if it'd work on earlier versions of Qt libs) :
libQt5Charts.so.5
libQt5Widgets.so.5
libQt5VirtualKeyboard.so.5
libQt5QuickControls2.so.5
libQt5Quick.so.5
libQt5Gui.so.5
libQt5QmlModels.so.5
libQt5Qml.so.5
libQt5WebSockets.so.5
libQt5Network.so.5
libQt5Core.so.5
- Depending on the system, you may need to install these libraries additionally:
libbcm_host.so
libvcos.so
libvchiq_arm.so
librt.so.1
libdl.so.2
libopenmaxil.so
libvcsm.so
libbrcmGLESv2.so
libpthread.so.0
libstdc++.so.6
libm.so.6
libgcc_s.so.1
libc.so.6
libbrcmEGL.so
libpng16.so.16
libz.so.1
libgthread-2.0.so.0
libglib-2.0.so.0
libpcre.so.3
Client/python ⇾ v1.2.0
- https://pypi.org/project/iot-facerecognition-client/1.2.0
- Requires 'opencv-python', 'websocket-client', 'screeninfo'.
- Support is deprecated in favor of qt-based client.
Minor Update
Server: v1.1.0, Client: v1.1.8
Server is built with Qt 5.15.0 & MSVC 2019. Qt and OpenCV libraries are included.
- Requires latest Visual C++ runtimes! VC++ Runtimes are NOT included. More info: https://support.microsoft.com/en-us/help/2977003/the-latest-supported-visual-c-downloads
- CUDA enabled version requires CUDA toolkit & CUDNN installation: https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html
Client can also be installed by "pip install iot-facerecognition-client".
Initial release - v1.0.0
Server: Built with Qt 5.14.2 & MSVC 2019 x64. Qt and OpenCV libraries are included.
Server: Requires the latest MSVC runtime to work!