Need inspiration for your next Edge AI model? Looking to skip some of the tedious data collection and labeling when starting your next project? Imagimob offers you DEEPCRAFT™ Studio Accelerators (formerly called 'Starter Models') for free to give you a jump start in your machine learning development!
DEEPCRAFT™ Studio Accelerators are open-source models that include the datasets, preprocessing, model architecture and instructions that developers need to turn them into production-ready Edge AI models. Studio Accelerators cover a wide variety of use cases and sensors, and are ready to be refined and trained based on your needs. The purpose is to inspire developers and to help kick-start your Edge AI journey.
If you have DEEPCRAFT™ Studio installed, you already have access to DEEPCRAFT™ Studio Accelerators. Studio comes with 3000 minutes of compute per month and is free for development, evaluation and testing, so there's nothing to hold you back from getting started now!
This machine learning project contains everything you need to develop and deploy your very own baby cry detection model. Bundled with the project is an already trained model, and instructions for how to deploy it on the Infineon PSOC™ 6, which is ideally suitable for consumer applications.
This DEEPCRAFT™ Studio Accelerator detects gunshots in a noisy environment. The model includes strong invariance to many different background noises, and has around one hour of microphone data. A limitation with this model is difficulty in testing.
This DEEPCRAFT™ Studio Accelerator classifies if there is a chainsaw actively cutting material in the vicinity. Chainsaws that are stalling are defined as not cutting. A fully developed model based on this Studio Accelerator could be used to detect illegal logging or to create automatic warning systems.
This DEEPCRAFT™ Studio Accelerator is capable of detecting a number of audio signatures common to a home setting. It currently has three labels; 'cough', 'baby cry', and 'water tap'. This Studio Accelerator can easily be modified to add more labels. This contains 550 minutes of data, most of which is unlabeled background noise.
This DEEPCRAFT™ Studio Accelerator is capable of classifying the material a power drill is drilling into based on the audio signature. It is designed to be incorporated into smart power tools. It is developed as a proof of concept and is not fully optimized, achieving around a 85% plastic/wood accuracy. It differentiates between wood, plastic and air.
This DEEPCRAFT™ Studio Accelerator offers a framework to develop a surface detection project for recognizing different surfaces (Floor, Carpet, or Air) based on sound patterns generated by a vacuum cleaner. The goal is to enable classification of these surfaces by analyzing the vacuum cleaner's audio signals as it operates. This is a classification project, a type of Supervised Learning where the system categorizes data into distinct classes. In this project, the three classes are Floor, Carpet, and Air.
This machine learning project contains everything you need to develop and deploy your very own siren detection model. Bundled with the project is an already trained model, and instructions for how to deploy it to the Infineon AURIX™ TC375 Lite Kit Board and KITA2G Audio Shield Board, which is ideally suitable for automotive and industrial applications.
This machine learning project contains everything to get started with keyword detection. Bundled with the project is a trained model, the Google Speech commands dataset, and the guide for how to download and prepare the dataset. You also get some hints on how to take the model to production.
This DEEPCRAFT™ Studio Accelerator allows you to build models to detect a fall using an IMU (accelerometer and gyroscope) mounted on the buckle of a belt. For that, this Accelerator Model uses data collected from two different IMUs: a Bosh IMU and an STMicroelectronics IMU. Both sensors are set up to collect data at 50 Hz using a +- 8g for the accelerometer scale and +- 500 dps for the gyro scale.
This DEEPCRAFT™ Studio Accelerator allows you to build a human activity detector that can be used on any supported Infineon MCU (or other MCUs) with a BMI160 IMU or another IMU. You can use this project as a starting point to develop a production-ready model intended for deployment in wrist-worn wearables.
This DEEPCRAFT™ Studio Accelerator is capable of classifying the material a power drill is drilling into based on the IMU (6 axis accelerometer and gyroscope) signature. It is designed to be incorporated into smart power tools. It differentiates between wood, plastic and air.
This DEEPCRAFT™ Studio Accelerator aims to provide general guidance on how to develop an anomaly detection system for detecting anomalous behavior in machinery based on vibration measurements. This project will monitor a simple desktop fan, but the same concept and workflow can be easily ported to any other machinery, whether industrial or consumer.
This is a simple DEEPCRAFT™ Studio Accelerator capable of differentiating between 3 different movement types: circle, shaking and stationary based on the IMU (6-axis accelerometer and gyroscope) of the AI Evaluation Kit. Note that in this project, stationary is unlabelled. This project serves as a code example but can also be adapted and expanded if you have an interesting application.
This DEEPCRAFT™ Studio Accelerator allows you to build a touch detection model that can be used on any supported Infineon MCU with CAPSENSE™. This Studio Accelerator gives you the infrastructure you need to expand on the project or to mimic it and create your own project based on the available/included data and tools.
This DEEPCRAFT™ Studio Accelerator is a real-time gesture recognition model powered by YOLO-based backend for detecting and classifying hand gestures. The project aims to build a robust end-to-end system that identifies rock, scissors, and paper gestures from live video input.
This DEEPCRAFT™ Studio Accelerator allows you to build a model that can detect termites. The model aims to build a robust end-to-end system that identifies termites from live video input. The YOLO-based termite object detection model can be applied in various fields, such as biology research for termite population and species counting, as well as in house and garden maintenance to manage termite infestations.
Want to skip the machine learning development altogether? Our machine learning engineers have made DEEPCRAFT™ Ready Models for you! These are off-the-shelf, fully-developed AI models ready for you to deploy in your products.
We are continuously releasing new DEEPCRAFT™ Ready Models, with various sensors and uses. Our Ready Models are based on:
- Audio: for detecting specific sounds
- Radar: for detecting presence, objects or materials
- IMU/Accelerometer: for detecting the motion of humans such as falling or vibrations in a machine
- Complete models ready to be put into production
- Fast time to market - no development or AI know-how required
- High accuracy and production quality
- Fully tested on a large library of datasets to ensure model works in wide range of scenarios
- Low memory footprint and inference time
- Data is kept local and not shared to the cloud
- Can be launched right away on embedded hardware (deployment-ready Infineon MCUs such as PSOC™ 6, PSOC™ Edge, AURIX™ and TRAVEO™. Talk to us about deploying on other hardware).
- Easy to integrate and deploy into consumer products
- Using ML models to add new features makes them future-proof, model improvements and updates with OTA software in ModusToolbox™
- Designed to work with wearables, health and safety devices that are familiar and convenient for end-users
Interested in talking to us about something our Edge AI tools, products and solutions? Get in touch so we can help you get AI features into your device as soon as possible!