The most popular and comprehensive Open Source ECM platform
Let us know more
GLOM: Attempting to Make Computer Vision Better by Mimicking How Humans See
Human brains are set up to be able to intuit from partial views of objects information about the size, shape and type of thing that is being viewed. And it is able to do it quickly — in fractions of seconds. This intuitiveness is something that computer vision algorithms are lacking. Vast amount of computing are currently needed to be able to match some the simple vision capabilities that come intuitively to humans.
One of the pioneers of Artificial Intelligence, Geoffrey Hinton, has an idea for how to make computer vision more similar to how humans experience vision. Hinton calls the new method GLOM (named after the slang term of ‘glomming things together’, or agglomeration), an idea inspired from mathematics and biology, and which attempts to make neural networks smarter.
Hinton said that “for humans, vision is really a sampling process, where the eye makes real time decisions around what information in the field of vision is going to be further deciphered. For example, we’re very good at quickly sampling and processing anything that moves. The same is true for something that has a different color from its background. When the eye fixates, whatever is in the middle of the retina is at high resolution and whatever is around the edge of that will be at a low resolution. You process what’s in focus for several 100 milliseconds before your eye fixates on something else. GLOM is about the deep learning process that happens after the system fixates on an image. It addresses a research problem I’ve had for the past 50 years and feels much closer to an understanding of how the brain might be doing vision… With GLOM, you should be able to recognize images based on the relationships between parts of an image, not on fine textures and things that aren’t discernible to the human eye.”
