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 Workflows
- Insight into the stages of the TinyML lifecycle
- Attributes of edge hardware components
- Strategic considerations for workflow design
Acquiring and Refining Data
- Collecting both structured datasets and raw sensor inputs
- Strategies for data annotation and augmentation
- Tailoring datasets for resource-constrained environments
Developing Models for TinyML
- Choosing appropriate model architectures for microcontrollers
- Establishing training processes with mainstream ML frameworks
- Assessing key model performance metrics
Optimizing and Compressing Models
- Applying quantization methods
- Utilizing pruning and weight-sharing techniques
- Striking a balance between precision and resource availability
Converting and Packaging Models
- Exporting models to TensorFlow Lite
- Embedding models within specialized toolchains
- Addressing model size and memory limitations
Implementing on Microcontrollers
- Writing models to physical hardware targets
- Setting up run-time execution environments
- Conducting real-time inference assessments
Oversight, Testing, and Verification
- Approaches for testing deployed TinyML systems
- Diagnosing model behavior on physical hardware
- Validating performance under field conditions
Assembling the Complete End-to-End Workflow
- Creating automated operational pipelines
- Managing versions of data, models, and firmware
- Overseeing updates and continuous iterations
Wrap-Up and Future Directions
Requirements
- A solid grasp of core machine learning principles
- Proficiency in embedded systems programming
- Comfort with Python-driven data processing workflows
Intended Learners
- AI specialists
- Software engineers
- Embedded systems architects
