This course teaches learners how to achieve real-time inference performance in intelligent video analytics by leveraging NVIDIA's approach, including tuning AI models for performance and optimizing GPU usage. The course covers topics such as transfer learning, data augmentation, automatic mixed precision, quantization, network pruning, and kernel auto-tuning. The teaching method includes demonstrations, discussions on software tools like DeepStream SDK, TensorRT, and Triton Inference Server, and insights on NVIDIA's end-to-end AI workflow. The intended audience for this course includes data scientists, computer vision professionals, and individuals interested in real-time intelligent video analytics.
Best Practices for Real-time Intelligent Video Analytics
Overview
Syllabus
Intro
Why intelligent video analytics?
Challenges with intelligent video analytics
Tips & tricks for efficient AI video analytics
Transfer learning
Data augmentation
Automatic mixed precision AMP
Quantization
Network pruning
Network graph optimizations
Kernel auto-tuning
Dynamic tensor memory upon inference
Multistream concurent execution
Free NVIDIA products designed to make your AI App efficient
NVIDIA's end-to-end AI workflow
TAO toolkit
high performance pre-trained vision AI models
Enabling beyond pre-trained AI models
Achieving state of the art accuracy for public datasets
NVIDIA TensorRT
TensorRT workflow
Triton inference server
DeepStream SDK
DeepStream application architecture
Pipelin efficiency with zero memory copies
NVIDIA graph composer
DeepStream video demo
Summary
Developer resources
Outro
Taught by
GOTO Conferences
