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 Ascend 310
- Edge AI overview: trends, constraints, and applications
- Architecture of the Huawei Ascend 310 chip and its supported toolchain
- The role of CANN within the edge AI deployment stack
Model Preparation and Transformation
- Exporting trained models from TensorFlow, PyTorch, and MindSpore
- Utilizing ATC to transform models into OM format for Ascend devices
- Addressing unsupported operations and lightweight transformation strategies
Creating Inference Pipelines with AscendCL
- Leveraging the AscendCL API to execute OM models on Ascend 310
- Input/output preprocessing, memory management, and device control
- Releasing within embedded containers or lightweight runtime environments
Optimizing for Edge Limitations
- Minimizing model size and precision adjustment (FP16, INT8)
- Identifying bottlenecks using the CANN profiler
- Managing memory layout and data streaming to boost performance
Releasing with MindSpore Lite
- Employing the MindSpore Lite runtime for mobile and embedded targets
- Comparing MindSpore Lite against raw AscendCL pipelines
- Packaging inference models for device-specific release
Edge Deployment Scenarios and Case Studies
- Case study: smart camera featuring an object detection model on Ascend 310
- Case study: real-time classification in an IoT sensor hub
- Monitoring and updating deployed models at the edge
Summary and Future Directions
Requirements
- Experience with AI model development or release workflows
- Foundational understanding of embedded systems, Linux, and Python
- Familiarity with deep learning frameworks like TensorFlow or PyTorch
Target Audience
- IoT solution developers
- Embedded AI engineers
- Edge system integrators and AI release specialists
Testimonials (1)
That we can cover advance topic and work with real-life example
