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
Overview of CV/NLP Deployment with CANN
- The AI model lifecycle from training through to deployment.
- Critical performance factors for real-time CV and NLP applications.
- An introduction to CANN SDK tools and their function in model integration.
Preparation of CV and NLP Models
- Exporting models from PyTorch, TensorFlow, and MindSpore.
- Managing model inputs and outputs for image and text-based tasks.
- Converting models to OM format using ATC.
Implementing Inference Pipelines with AscendCL
- Executing CV/NLP inference via the AscendCL API.
- Preprocessing steps: image resizing, tokenization, and normalization.
- Postprocessing tasks: generating bounding boxes, classification scores, and text output.
Techniques for Performance Optimization
- Profiling CV and NLP models with CANN tools.
- Minimizing latency through mixed-precision and batch tuning.
- Efficient management of memory and compute resources for streaming tasks.
Computer Vision Applications
- Case study: object detection in smart surveillance systems.
- Case study: visual quality inspection in manufacturing processes.
- Developing live video analytics pipelines on Ascend 310.
NLP Applications
- Case study: sentiment analysis and intent detection.
- Case study: document classification and summarization.
- Integrating real-time NLP services with REST APIs and messaging systems.
Conclusion and Future Directions
Requirements
- Foundational knowledge of deep learning in computer vision or NLP.
- Hands-on experience with Python and AI frameworks like TensorFlow, PyTorch, or MindSpore.
- A basic grasp of model deployment and inference workflows.
Target Audience
- Computer vision and NLP professionals utilizing Huawei’s Ascend platform.
- Data scientists and AI engineers working on real-time perception models.
- Developers integrating CANN pipelines into manufacturing, surveillance, or media analytics systems.
