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
TinyML Pipeline Fundamentals
- Summary of TinyML workflow phases
- Attributes of edge hardware
- Key aspects in pipeline architecture
Data Acquisition and Preprocessing
- Gathering structured and sensor-based data
- Approaches for data labeling and augmentation
- Preparing datasets for restricted environments
Model Construction for TinyML
- Choosing model structures for microcontrollers
- Training procedures using mainstream ML frameworks
- Assessing model performance metrics
Model Refinement and Compression
- Quantization methods
- Pruning and weight sharing techniques
- Striking a balance between accuracy and resource limitations
Model Transformation and Packaging
- Exporting models to TensorFlow Lite
- Incorporating models into embedded toolchains
- Controlling model size and memory usage
Deployment on Microcontrollers
- Installing models onto hardware targets
- Setting up run-time environments
- Conducting real-time inference tests
Monitoring, Testing, and Validation
- Testing methodologies for deployed TinyML systems
- Troubleshooting model behavior on hardware
- Validating performance in field conditions
Implementing the Complete End-to-End Pipeline
- Creating automated workflows
- Versioning data, models, and firmware
- Overseeing updates and iterative improvements
Recap and Future Directions
Requirements
- A solid grasp of core machine learning principles
- Practical experience in embedded programming
- Proficiency with Python-centric data workflows
Target Audience
- AI engineers
- Software developers
- Experts in embedded systems
