Resource constraints and safety verification for embedded AI/ML

Overview

The talk examines memory/compute constraints and pruning or quantization of embedded AI models. It introduces combining model-change control, protective logic, traceability and testing for safety-related applications including automotive systems.