Edge and Embedded AI
This video series introduces edge and embedded AI, where trained models run directly on local devices, such as edge computers or embedded systems, rather than relying on the cloud. Learn how this approach enables faster responses, reduces bandwidth usage, enhances data privacy, and improves reliability even when network connections are unstable.
Discover how developing AI for edge and embedded systems differs from cloud deployment. While the fundamental workflow is similar, extra attention must be given to optimizing models for devices with limited resources. The videos cover practical strategies such as pruning, quantization, and projection to help your models fit and perform efficiently on target hardware. Through hands-on examples, see how neural network models can be compressed—sometimes by over 90%—without sacrificing accuracy, making it possible to bring advanced AI capabilities to resource-constrained devices.
A Practical Introduction to Edge AI
Learn about the practical challenges of edge AI, including validation, optimization, deployment, and reliability considerations.
Compressing Neural Networks for Embedded AI: Pruning, Projection, and Quantization
Learn how to compress neural networks using pruning, projection, and quantization to run efficiently on embedded devices—without losing accuracy.
