Become A DSP Tuning Master and Build More Efficient Neural Networks

When: 2021-10-06 13:00 EDT
Add to Calendar
Duration: 1h00
Instructor: Alex Elium
Description:

Sensor data is typically preprocessed with DSP in TinyML applications.  As engineers deploy NNs on ever smaller processors, it is becoming necessary to tune DSP algorithms in order to fit within RAM or real-time processing constraints.  But not all steps in a DSP pipeline are created equal!  Knowing how to find sections to slim down can mean the difference between giving up a few percent of accuracy, and ending up with a model that’s no longer usable.

This presentation will show experimentation with DSP parameter choices (number of cepstral coefficients, spectrogram frame size, etc) for an example keyword spotting classifier, and analyze the RAM, latency, and accuracy impacts of various scenarios.  Attendees will leave with ideas on where to find elusive kB of RAM and mS of latency next time they need to optimize a DSP pipeline.

Alex Elium is an embedded software engineer at Edge Impulse, where he focuses on DSP for image classifiers, keyword spotting, and other TinyML networks. He also works on integration of NN hardware acceleration frameworks, including NVIDIA's TensorRT and ARM's Vela compiler for NPUs. In his career, he has developed numerous embedded systems, including those for acoustic localization of beacons, acoustic digital communication, and a Doppler velocity log for navigation underwater.

Edge Impulse is the leading development platform for machine learning on edge devices, free for developers and trusted by enterprises.