We are interfacing with the real world, developing algorithms to process real data, and delivering unique processing solutions.
We collect large-scale real world data to develop machine learning, statistical inference, deep learning, and signal processing algorithms.
We create new sensors that we can use in our advanced algorithms to deliver better solutions.
We invent embedded hardware to run our algorithms in computationally-efficient and reasonable power envelopes, enabling advanced algorithms to be used where they could not before. We build solutions that are flexible - taking advantage of cloud compute resources when the situation dictates or embedded hardware where necessary.
We develop digital predistortion technology to make cellular basestations deliver better signals to your smartphone at longer distances.
We are developing microprocessors that will run machine learning, statistical inference, and deep learning algorithms in computationally efficient and reasonable power envelopes.
We are deeply interested in improving the accuracy of spatial measurement across many types of sensing modalities: radar, LIDAR, and depth imaging to enable cars, drones, and robots to get around our world more effectively.
We are researching methods to measure real world biological signals non-invasively, in the presence of significant noise. We leverage the tools of signal processing and machine learning to do this effectively.
We are developing advanced algorithms to enable voice-based user interfaces to work better in noisy environments or situations where the device is far from the user. We are developing both cloud-hosted services as well as embedded hardware implementations to enable people to engage our technology regardless of their product development constraints.