Bookmark and Share


Google has built two generations of large-scale systems for training neural networks, applying these systems to a wide variety of problems.

This column highlights already-published content from previous Summits and other Alliance events, related to themes covered in past columns.

On-phone image processing is a recent phenomenon, enabled by modern devices' increasingly robust hardware, software and networks.

Autonomous driving is now being directly addressed by every major automobile manufacturer, Tier 1 and a number of new entrants in the field.

In time, with further vision research and development attention, memory and processing requirements will inevitably continue to diminish.

Vision processing is a necessity for robust autonomy.

Deploying autonomous cars widely is going to take a while. In the meantime, autonomous vehicles of other kinds are already on sale.

Practical computer vision processing is becoming increasingly feasible at both low power and low price points.

ADAS (advanced driver assistance systems) is rapidly becoming a huge vision processing technology success story.

"Deep learning" techniques such as convolutional neural networks offer compelling advantages over conventional computer vision algorithms.