Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore simplified AI solutions for embedded applications in this 57-minute talk. Discover how to address the constraints of embedded systems by reducing the complexity of deep net inference engines. Learn techniques for minimizing intra-network connectivity, eliminating floating-point data requirements, and replacing multiply-accumulate operations with simple accumulation. Gain insights into developing small-footprint, low-latency deep nets suitable for various embedded applications using 8/16/32/64-bit MCUs, DSPs, and FPGAs. Understand the potential applications of these solutions in IoT smart sensors for inertial, vibration, temperature, flow, electrical, and biochemical measurements in battery-powered endpoints, including healthcare/industrial wearables, robots, and automotive systems.