US-based Ambarella is collaborating with China-based Momenta to introduce a HD mapping platform for self-driving vehicles.

The platform has been designed using Ambarella’s CV22AQ CVflow computer vision system-on-chip (SoC) along with Momenta’s deep learning algorithms.

The collaborative HD mapping solution offers mapping and localisation for autonomous vehicles, as well as map updates through crowdsourcing.

Momenta CEO Xudong Cao said: “HD map is essential to autonomous driving systems. Ambarella’s CV22AQ CVflow computing platform has made it easier for Momenta to deploy and upgrade our HD map software and algorithm on embedded systems.

“Our HD map can automatically build maps, perform localisation, and update through crowdsourcing.”

Through crowdsourcing, Momenta’s vision-based HD semantic mapping solution can generate a closed feedback loop of big data, AI, and HD map updates.

“HD map is essential to autonomous driving systems.”

To spot the changes in the map elements, Momenta uses localisation technology and then delivers frequent updates to the Cloud.

The CV22AQ has been designed to consume very little power and is capable of real-time processing of up to eight megapixel resolution video at 30 frames per second (fps) for high-precision deep-learning based object recognition.

Leveraging CV22AQ, Momenta uses a single monocular camera input to generate two separate video outputs.

It produces one for vision sensing with perception of lanes, traffic signs, and other objects. Another feature is point extraction for self-localisation and mapping (SLAM) and optical flow algorithms.

Ambarella CEO Fermi Wang said: “The CV22AQ’s performance and advanced image processing help enable the full potential of Momenta’s advanced AI algorithms.”

Ambarella offers a set of tools to allow users easily port their neural networks to the CV22AQ SoC.