AidLearning FrameWork
v1.3.0
AidLux 1.3 is officially released! (Watch the video ⬆️)
AI, Android, Linux, ARM: AI application development platform based on Android+Linux integrated ecosystem. [English]
[AidLux](https://docs.aidlux.com) is an AI application development and deployment platform built on ARM and supporting multi-ecological integration (Android+Linux) environment, bringing powerful, simple and unlimited creativity to developers. What an amazing experience possible! A complete Linux system environment is built by sharing the Android Linux kernel, and provides user access at the same time as the Android system environment. While providing users with a command line experience similar to native Linux systems (such as package management through the apt command), a graphical desktop environment is built based on the Web, which users can access directly through a touch screen or browser.
AidLux completes all basic scientific computing packages/libraries required for AI operation, supports the industry's mainstream deep learning frameworks, and has built-in self-developed AI intelligent acceleration technology, providing developers with an "AI-ready" application development platform.
A device runs two system environments at the same time. It is both an Android device and a Linux device. The resource advantages of the two ecologies can be utilized simultaneously.
Seamless interaction across systems, efficient, safe and stable. Android applications and Linux applications achieve local and efficient direct access without the need for external interfaces (such as networks), fully releasing hardware performance. Android is responsible for user interaction, and Linux is responsible for service support. New application forms are waiting for your whims.
For existing applications (Android or Linux), cross-system function support can be easily obtained, making the application more powerful and fully protecting existing development investments.
One-click installation, automatic deployment, and App-style startup make the process fast and simple.
Integrating the industry's mainstream AI deep learning framework, no configuration is required, and it can be installed and used immediately, which greatly reduces the complexity of environment configuration for AI development and application deployment, and significantly reduces related time investment. The framework supports the following:
| TensorFlow | PyTorch | Caffe | MXNet | Keras | MindSpore | PandlePandle | TNN | MNN | SNPE |
|---|---|---|---|---|---|---|---|---|---|
| √ | √ | √ | √ | √ | √ | √ | √ | √ | √ |
Built-in innovative CPU+GPU+NPU intelligent acceleration technology, through "hardware+framework+Op" multi-layer optimization, greatly improves deep learning computing performance. It also provides a unified API interface, which is convenient for developers to call and also supports automatic conversion of different AI framework models. The following tests are based on the same ARM device and were tested on Linux and AidLux. A single test was performed 20 times and the average time (vertical axis) was taken.
Develop drag-and-drop AI applications in Wizard, quickly empower AI components, and generate your first AI application in 1 minute!
It has built-in rich AI application cases and corresponding codes to facilitate developers to get started quickly.
AidLux cloud desktop system supports simultaneous access from PC, tablet, mobile phone and other screens anytime and anywhere, realizing your super mobile development platform. Quickly enter the cloud desktop address through the browser on your PC to access the AidLux desktop on your mobile phone!
Easily supports a large number of peripherals (network, USB, serial port,...), and the creative space is infinitely expanded.
Supports a variety of open source development tools such as VSCode, Jupyter notebook, and development languages such as Python, C/C++, Java, and JavaScript.
One-stop full-process support for development, testing, and deployment. AidLux focuses on efficiency, and you focus on creativity! The ApkBuild application can quickly package projects developed based on Python into APKs for release, making it convenient for users to deploy on other Android systems.
Click the link below to download the latest installation package [AidLux v1.3.0]
AidLux participants: bill, flay, gondon, willam, gugu, yoline777, qidiso, yuge, muzi_ys, etc.
The following items:
