Case studies

Embedded DNN

PROJECT

Embedded DNN

INDUSTRY

Consumer Electronics

SERVICE

Convolutional Neural Net


The challenge was to create a
Convolutional Neural Net that would run on a constrained system.

PROCESS

The process was to first prototype a CNN running in real-time performing image recognition on a desktop class device. From there optimize the architecture, primarily the models so that they could run on an embedded device.

RESULT

Cross platform library that runs in real-time at greater then 25fps that performs image recognition on known datasets at >90% accuracy.

TECHNICAL DETAIL

TensorRT, C++
Initially running on ARM Cortex A-series
Currently working on M-series

Clients and Customers First

Great ideas come from collaboration.