Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Duration 21 hours
Course Outline
Core Principles of TinyML Pipelines
- Understanding the stages of the TinyML workflow
- Key attributes of edge hardware
- Strategic considerations for pipeline design
Acquisition and Preprocessing of Data
- Gathering structured and sensor-based data
- Strategies for data labeling and augmentation
- Curating datasets suitable for constrained environments
Building Models for TinyML
- Choosing model architectures compatible with microcontrollers
- Training processes leveraging standard ML frameworks
- Assessing key performance metrics
Optimizing and Compressing Models
- Methods for quantization
- Pruning and weight sharing techniques
- Achieving balance between accuracy and resource limitations
Model Transformation and Packaging
- Exporting models to TensorFlow Lite
- Embedding models within toolchains
- Handling model size and memory restrictions
Implementation on Microcontrollers
- Programming models onto hardware targets
- Setting up runtime environments
- Conducting real-time inference tests
Overseeing, Testing, and Verification
- Testing methodologies for deployed TinyML systems
- Troubleshooting model performance on hardware
- Validating performance in operational field conditions
Assembling the Complete End-to-End Workflow
- Creating automated processing streams
- Managing versions of data, models, and firmware
- Coordinating updates and iterative improvements
Recap and Future Directions
Requirements
- A solid grasp of machine learning fundamentals
- Practical experience in embedded programming
- Comfort with Python-based data processing workflows
Target Audience
- AI Engineers
- Software Developers
- Embedded Systems Specialists
