Description: Deep neural networks are revolutionizing machine perception, bringing incredible new capabilities to many types of systems. But developing DNNs for edge devices is typically a time-consuming and error-prone process. In this session:

• We’ll show how the NetsPresso AI model optimization platform drastically simplifies the process of selecting a model, training it, compressing it and deploying it — taking into account the specific capabilities and limitations of the target hardware.
• We’ll show how NetsPresso uses neural architecture search to quickly find the best model for your specific application and hardware, and then trains the model in a hardware-aware manner to optimize accuracy and latency for your processor.
• Next, we’ll explain how NetsPresso automatically applies model compression and acceleration techniques to make your model small and fast without sacrificing accuracy.
• Finally, we’ll show how NetsPresso simplifies deployment of optimized models on embedded hardware by automatically generating executable code and packaging it in a form that can easily be integrated into your application.

We’ll illustrate these capabilities using real-world use cases and applications, and we’ll evaluate the optimized models produced by NetsPresso.
Join us to learn how you can create, optimize and deploy DNNs at the edge in days rather than months.