Questions? Answers!
We answer your questions about the AI framework and Learn AIfES. If you want to know even more, don't be afraid to keep asking our experts here on the wall or on LinkedIn.
What do I need to get started with AIfES?
Since AIfES is open source and therefore accessible to everyone for free, you can basically get started right away! For an easy start we would like to offer you free webinars with Learn AIfES. No matter if you are already a professional or a newbie in the field of machine learning, we want to teach you AIfES and turn you into real AI experts.
What devices does AIfES run on?
The great thing about AIfES is that it can be used on almost any system, whether it's a microcontroller, IoT device, Raspberry PI, PC or smartphone. So you can skip buying new hardware and get started right away!
For which ML problems is AIfES suitable?
Basically for any machine learning problem that you can solve with complex neural networks. You still need inspiration? Our demonstrators can help: gesture recognition, color and object recognition, but also an interactive tic-tac-toe game plus the corresponding code.
Are you smarter than AIfES? Try it out:Tic-Tac-Toe Simulator.
Learn AIfES and quality education
.
The project supports the UN Sustainable Development Goal to advance AI for quality education. This goal aims to ensure inclusive, equitable, and quality education and promote lifelong learning opportunities for all.
Through your support, you will help the Learn AIfES team bring the free webinars and possible other cool formats and developments to the world. True to the open source idea, the results will be made freely available to all.
Learn AIfES and climate protection
Deep Learning on high performance computers can emit enormous amounts of CO2. In 2019, the University of Massachusetts created a life cycle assessment for training large AI models. For example, the process of natural language processing (NLP) can emit more than 283.9 tons of CO₂. That's nearly five times the lifetime emissions of an average American car. Because of this, experts are calling for more efficient algorithms and models, as well as hardware that uses less energy (Strubell et al. 2019).
As our project gives individuals, as well as industry, a framework to do exactly that, we are giving it our all to support the UN's Sustainable Development Goal of climate action.
What is so exciting about Tiny ML?
According to the TinyML Foundation's definition, Tiny Machine Learning (TinyML) is a rapidly growing field of machine learning technologies and applications. It encompasses hardware, algorithms, and software capable of analyzing sensor data on the sensor node at extremely low power (typically in the mW range and below), enabling a variety of applications even on battery-powered devices. Rapid progress is being made. For example, significant advances have been made in algorithms, networks, and models with a size of 100 kB and less, and the first low-power applications have been developed in the areas of image processing and audio.
AIFES is a pioneer and first enables data processing on embedded systems. This allows data to be stored on microcontrollers and small IoT devices and processing to be performed without transmission delay.
For more on TinyML, see theTinyML Foundation website.
How can I install AIfES
You can download and install AIfES by searching for "aifes" using the Arduino Library Manager. Alternatively, you can also download it manually. Download the AIfES repository as a ZIP archive and follow the instructions. You can find all information on our GitHub:
To what extent can the architecture be customized?
AIfES was designed as a flexible and extensible toolbox for running and training artificial neural networks on microcontrollers. All layers, loss and optimization functions are modular and can be optimized for different data types and hardware platforms. AIfES currently supports complex neural network types for inference and training.
By the way, the brand new Python AIfES Converter turns any TensorFlow or PyTorch model into an AIfES model with only two lines of Python code. The resulting AIfES model can then be used directly on your microcontroller.
References
- Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and policy considerations for deep learning in NLP. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 3645-3650, Florence, Italy. Association for Computational Linguistics.
- tinyML Foundation. Retrieved October 25, 2022, from https://www.tinyml.org/
