The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) AI at the Edge: Qualcomm AI Research at NeurIPS 2024 with Arash Behboodi - #711
21 snips
Dec 3, 2024 
Arash Behboodi, Director of Engineering at Qualcomm AI Research, discusses what's on the agenda for this year's NeurIPS conference. He highlights the challenges of differentiable simulation, particularly in wireless systems, and dives into how uncertainty quantification can enhance machine learning models through conformal prediction and entropy. Behboodi also previews innovative demos like on-device video editing and 3D content generation, showcasing Qualcomm's commitment to making cutting-edge AI accessible.
AI Snips
Chapters
Transcript
Episode notes
Simulators for Inverse Problems
- Many inverse problems involve complex forward processes, requiring simulators for various phenomena.
- Simulators help understand algorithm performance, like in wireless communication with electromagnetic waves.
Types of Wireless Simulators
- Wireless simulators range from statistical models to physically consistent ones based on ray tracing.
- These are used to solve various problems, including positioning, throughput, and latency.
Access Point Placement
- Finding the optimal access point location for best home coverage is an inverse problem example.
- Differentiable models are needed for optimization; otherwise, heuristics or Bayesian optimization are used.