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
Overview of TinyML and Embedded AI
- Key aspects of deploying TinyML models
- Limitations encountered in microcontroller settings
- Introduction to embedded AI development toolkits
Core Principles of Model Optimization
- Analyzing computational constraints
- Detecting operations with high memory demands
- Establishing baseline performance metrics
Quantization Methodologies
- Approaches to post-training quantization
- Implementation of quantization-aware training
- Balancing model accuracy against resource usage
Pruning and Data Compression
- Techniques for structured and unstructured pruning
- Leveraging weight sharing and model sparsity
- Applying compression algorithms for efficient inference
Optimization for Specific Hardware
- Model deployment on ARM Cortex-M architectures
- Enhancing performance via DSP and accelerator features
- Considerations for memory mapping and data flow
Performance Benchmarking and Verification
- Assessing latency and throughput
- Monitoring power and energy usage
- Conducting accuracy and robustness assessments
Deployment Processes and Tooling
- Utilizing TensorFlow Lite Micro for embedded implementations
- Connecting TinyML models with Edge Impulse workflows
- Debugging and testing on physical hardware
Advanced Refinement Tactics
- Applying neural architecture search to TinyML
- Combining quantization with pruning methods
- Using model distillation for embedded inference
Conclusion and Future Directions
Requirements
- A solid grasp of machine learning processes
- Background in embedded systems or microcontroller-based development
- Proficiency in Python programming
Target Audience
- AI researchers
- Engineers specializing in embedded ML
- Professionals developing inference systems under resource constraints
