Bookmark and Share


Key to an understanding of how computer vision will evolve is the reality that it's an enabling technology, not an end in itself.

"Deep learning" artificial neural networks are significantly better than previous techniques with a diversity of visual understanding tasks.

While employing vision and deep learning in embedded systems poses a challenge, it is becoming a requirement.

Whilst devices are everywhere you look, the financial gain is slower to catch up.

HoloLens has nailed both the "feels real" and ease-of-use aspects of a "mixed reality" glasses product.

Depth sensors' ability to discern an object's distance can make it feasible, or at least easier, to implement some vision functions.

Using unshielded twisted pair cable to deliver data, along with smaller and more compact connectors, can reduce connectivity cost up to 80%.

Why was SoftBank willing to pay so much more than the market thought the company was worth?

Computer vision isn't going to merely make cars better, it's going to completely transform the automotive industry.

The analogy with biological vision can help us figure out which computer vision capabilities and applications are worthwhile, vs novelties.