Designers looking to implement artificial intelligence (AI) and machine learning (ML) at the Edge, particularly for voice user interfaces (VUIs) and sound event detection (SED), need to find a way to balance cost, performance, and power, particularly for battery-powered devices. While executing AI/ML at the edge ensure privacy and low latency, two key contributors to a positive user experience, a poorly implemented, expensive, and unreliable design and can greatly impact that experience.

Attendees will come away with an understanding of:

  • Why Edge AI/ML is so critical and important to get right
  • What challenges designers face when implementing AI/ML
  • General architectural approaches toward addressing these challenges
  • Advances in algorithm compression techniques toward improved efficiency and performance
  • What to look for in a toolchain to ensure broad compatibility across popular AI frameworks, while allowing optimization and accelerated, cost-effective deployment