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.
Course Outline
Introduction to TinyML and Embedded AI
- Key characteristics of TinyML model deployment
- Constraints inherent to microcontroller environments
- Overview of embedded AI toolchains
Foundations of Model Optimisation
- Identifying computational bottlenecks
- Pinpointing memory-intensive operations
- Conducting baseline performance profiling
Quantisation Techniques
- Post-training quantisation strategies
- Quantisation-aware training
- Evaluating the balance between accuracy and resource usage
Pruning and Compression
- Structured and unstructured pruning methods
- Weight sharing and achieving model sparsity
- Compression algorithms designed for lightweight inference
Hardware-Aware Optimisation
- Deploying models on ARM Cortex-M systems
- Optimising for DSP and accelerator extensions
- Managing memory mapping and dataflow considerations
Benchmarking and Validation
- Analysing latency and throughput
- Measuring power and energy consumption
- Conducting accuracy and robustness testing
Deployment Workflows and Tools
- Leveraging TensorFlow Lite Micro for embedded deployment
- Integrating TinyML models with Edge Impulse pipelines
- Testing and debugging on actual hardware
Advanced Optimisation Strategies
- Neural architecture search for TinyML
- Hybrid quantisation-pruning approaches
- Model distillation for embedded inference
Summary and Next Steps
Requirements
- A solid understanding of machine learning workflows
- Experience in embedded systems or microcontroller-based development
- Proficiency in Python programming
Target Audience
- AI researchers
- Embedded ML engineers
- Professionals specialising in resource-constrained inference systems
21 Hours
