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
- Survey of TinyML workflow phases
- Attributes of edge computing hardware
- Strategic pipeline architecture factors
Acquisition and Preprocessing of Data
- Gathering structured information and sensor readings
- Tactics for data annotation and augmentation
- Structuring datasets for limited-resource contexts
Model Construction for TinyML
- Choosing architectures suited for microcontrollers
- Training procedures leveraging standard ML libraries
- Assessing key model performance metrics
Optimizing and Compressing Models
- Application of quantization methods
- Strategies for pruning and weight sharing
- Reconciling precision with hardware limitations
Conversion and Packaging of Models
- Exporting to TensorFlow Lite formats
- Incorporating models into embedded development environments
- Regulating model dimensions and memory usage
Implementation on Microcontrollers
- Writing models to physical hardware targets
- Setting up execution environments
- Conducting real-time inference assessments
Oversight, Verification, and Validation
- Approaches for testing deployed TinyML systems
- Troubleshooting model operations on physical devices
- Confirming performance under field conditions
Unifying the Complete End-to-End Pipeline
- Creating automated processing flows
- Managing versions for data, models, and firmware
- Overseeing updates and iterative improvements
Recap and Future Directions
Requirements
- Solid grasp of machine learning core concepts
- Practical experience in embedded coding
- Proficiency in Python-driven data handling processes
Target Audience
- AI specialists
- Software engineers
- Embedded systems architects
Testimonials (1)
The exercises