DEEPCRAFT™ Ready Models
DEEPCRAFT™ Ready Models are production-ready machine learning models for edge devices. Each model is delivered as a complete DEEPCRAFT™ Studio project, including data, preprocessor, trained model and Graph UX assets. You can evaluate the model as shipped or adapt the ready model to your product. This approach avoids the time and cost of collecting, labeling, and training a model from scratch. Use DEEPCRAFT™ Studio for the rest of the workflow when you need additional data, retraining, generated code, or on-device deployment. To know more, refer to our official website .
How to obtain the DEEPCRAFT™ Ready Models?
Ready Models are no longer requested through our website form and delivered by email. DEEPCRAFT™ Studio now delivers each Ready Model as a Studio project: create a new project, select the Ready Model you need, and Studio downloads the complete project into your workspace. To know more, refer to Getting started for detailed steps.
You can evaluate and deploy the Ready Model as shipped, or customize the Ready Model for your product. Earlier releases did not support customization; that restriction has been removed, so you work with the Ready Model as you would with any Studio project: collect more data, retrain, evaluate, generate code, and deploy. See the DEEPCRAFT™ Studio documentation for the full machine learning workflow.
Each Ready Model is shipped with performance figures and operating recommendations that apply only to the Ready Model as downloaded in Studio. Retraining the model or changing the architecture or parameters invalidates those figures. See the Ready Model Reports for the performance metrics.
Choose your Ready Model
The following audio ready models are available in DEEPCRAFT™ Studio:
| Model | Typical use | Optimal range | Target Application |
|---|---|---|---|
| Baby Cry Detection | Wearables and baby products; 0–4 years | Up to 5 m | Home Health Monitoring, Audio and Speaker - Home or Professional |
| Cough Detection | Health wearables and nearby devices | Up to 2 m | Home Health Monitoring, Audio and Speaker - Home or Professional |
| Snore Detection | Sleep wearables and near-bed devices | Up to 2 m | Home Health Monitoring, Audio and Speaker - Home or Professional |
| Siren Detection | Wearables, hearables, in-cabin alerts | Up to 20 m | Home Health Monitoring, Audio and Speaker - Home or Professional |
| Gesture Classification | Touchless control of a monitor or connected device | 10–70 cm, within 10 degrees | Collaborative Robot (Cobot), Commercial Building Automation, Commercial, Residential and Industrial Lighting, Domestic Robot (vacuum, lawn, pool), Drone & Multicopter, Humanoid Robots |
Refer to Ready Models for operating recommendations and datasets. Refer to Ready Model Reports for evaluation results.
Choose your Infineon AI Board
Ready Models are designed for specific boards to deliver the best performance. For evaluation, use the following kits.
What’s next?
To download a Ready Model as a Studio project, refer to Getting started.