In this week’s Embedded Insiders, Brandon and Rich try to decide if data sheets specs are reliable, or if industry benchmarks are the only reasonably- accurate measure of component performance without actually testing them yourself.
Classic DSP, New Neural Networks & Better Benchmarks Improve Local Voice Activation at the Edge - StoryJuly 19, 2021
If you’ve ever used a virtual assistant, you likely assumed you were talking to a device so smart it could answer almost any question you asked it. Well, actually, Amazon Echos, Google Homes, and other devices like them usually have no idea what you’re talking about.
MLCommons released new results for MLPerf Training v1.0, the organization's machine learning training performance benchmark suite.
MLCommons launched a new benchmark, MLPerf Tiny Inference, to measure how a trained neural network can process new data for low-power devices in small form factors and included an optional power measurement.