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 14 hours
Course Outline
Introduction to Edge AI and Model Optimization
- Grasping the nature of edge computing and AI workloads.
- Balancing performance against resource limitations.
- A look at various model optimization strategies.
Model Selection and Pre-training
- Selecting lightweight architectures (e.g., MobileNet, TinyML, SqueezeNet).
- Analyzing model structures fit for edge devices.
- Leveraging pre-trained models as a starting point.
Fine-Tuning and Transfer Learning
- Core principles of transfer learning.
- Adjusting models to fit custom datasets.
- Practical workflows for fine-tuning.
Model Quantization
- Techniques for post-training quantization.
- Quantization-aware training methods.
- Assessing performance trade-offs.
Model Pruning and Compression
- Pruning approaches (structured versus unstructured).
- Compression methods and weight sharing.
- Benchmarking the efficiency of compressed models.
Deployment Frameworks and Tools
- Using TensorFlow Lite, PyTorch Mobile, and ONNX.
- Ensuring edge hardware compatibility and managing runtime environments.
- Toolchains facilitating cross-platform deployment.
Hands-On Deployment
- Implementation on Raspberry Pi, Jetson Nano, and mobile devices.
- Performance profiling and benchmarking.
- Resolving common deployment challenges.
Summary and Future Directions
Requirements
- A solid grasp of fundamental machine learning concepts.
- Proficiency in Python and experience with deep learning frameworks.
- Knowledge of embedded systems and the constraints of edge devices.
Intended Audience
- Developers specializing in embedded AI.
- Experts in edge computing.
- Machine learning engineers with a focus on edge deployment.