Chris Rowen, Fellow at Cadence, presents the "Taming the Beast: Performance and Energy Optimization Across Embedded Feature Detection and Tracking" tutorial at the May 2014 Embedded Vision Summit.
This presentation looks at a cross-section of advanced feature detectors, and considers the algorithm, bit precision, arithmetic primitives and implementation optimizations that yield high pixel processing rates, high result quality and low energy. Rowen also examines how these optimization methods apply to kernels used in tracking applications, including fast connected component labeling. From this Rowen derives general principles on the priority and likely impact of different optimization types.
