Computer vision algorithms have come a long way in the past decade. They've been shown to be as good or better than people at tasks like categorizing dog or cat breeds, and they have the remarkable ability to identify specific faces out of a sea of millions.
But research by Brown University scientists shows that computers fail miserably at a class of tasks that even young children have no problem with: determining whether two objects in an image are the same or different. In a paper presented last week at the annual meeting of the Cognitive Science Society, the Brown team sheds light on why computers are so bad at these types of tasks and suggests avenues toward smarter computer vision systems.
Computers are great at categorizing images by the objects found with them, but they're surprisingly bad at figuring out when two objects in a single image are the same or different from each other. New research helps to show why that task is so difficult for modern computer vision algorithms.
Credit: Serre lab / Brown University
"There's a lot of excitement about what computer vision has been able to achieve, and I share a lot of that," said Thomas Serre, associate professor of cognitive, linguistic and psychological sciences at Brown and the paper's senior author. "But we think that by working to understand the limitations of current computer vision systems as we've done here, we can really move toward new, much more advanced systems rather than simply tweaking the systems we already have."
For the study, Serre and his colleagues used state-of-the-art computer vision algorithms to analyze simple black-and-white images containing two or more randomly generated shapes. In some cases the objects were identical; sometimes they were the same but with one object rotated in relation to the other; sometimes the objects were completely different. The computer was asked to identify the same-or-different relationship.
The study showed that, even after hundreds of thousands of training examples, the algorithms were no better than chance at recognizing the appropriate relationship. The question, then, was why these systems are so bad at this task.
Serre and his colleagues had a suspicion that it has something to do with the inability of these computer vision algorithms to individuate objects. When computers look at an image, they can't actually tell where one object in the image stops and the background, or another object, begins. They just see a collection of pixels that have similar patterns to collections of pixels they've learned to associate with certain labels. That works fine for identification or categorization problems, but falls apart when trying to compare two objects.
To show that this was indeed why the algorithms were breaking down, Serre and his team performed experiments that relieved the computer from having to individuate objects on its own. Instead of showing the computer two objects in the same image, the researchers showed the computer the objects one at a time in separate images. The experiments showed that the algorithms had no problem learning same-or-different relationship as long as they didn't have to view the two objects in the same image.
The source of the problem in individuating objects, Serre says, is the architecture of the machine learning systems that power the algorithms. The algorithms use convolutional neural networks -- layers of connected processing units that loosely mimic networks of neurons in the brain. A key difference from the brain is that the artificial networks are exclusively "feed-forward" -- meaning information has a one-way flow through the layers of the network. That's not how the visual system in humans works, according to Serre.
"If you look at the anatomy of our own visual system, you find that there are a lot of recurring connections, where the information goes from a higher visual area to a lower visual area and back through," Serre said.
While it's not clear exactly what those feedbacks do, Serre says, it's likely that they have something to do with our ability to pay attention to certain parts of our visual field and make mental representations of objects in our minds.
"Presumably people attend to one object, building a feature representation that is bound to that object in their working memory," Serre said. "Then they shift their attention to another object. When both objects are represented in working memory, your visual system is able to make comparisons like same-or-different."
Serre and his colleagues hypothesize that the reason computers can't do anything like that is because feed-forward neural networks don't allow for the kind of recurrent processing required for this individuation and mental representation of objects. It could be, Serre says, that making computer vision smarter will require neural networks that more closely approximate the recurrent nature of human visual processing.
Serre's co-authors on the paper were Junkyung Kim and Matthew Ricci. The research was supported by the National Science Foundation (IIS-1252951, 1644760) and DARPA (YFA N66001-14-1-4037).
Kevin Stacey | EurekAlert!
New Foldable Drone Flies through Narrow Holes in Rescue Missions
12.12.2018 | Universität Zürich
NIST's antenna evaluation method could help boost 5G network capacity and cut costs
11.12.2018 | National Institute of Standards and Technology (NIST)
The more objects we make "smart," from watches to entire buildings, the greater the need for these devices to store and retrieve massive amounts of data quickly without consuming too much power.
Millions of new memory cells could be part of a computer chip and provide that speed and energy savings, thanks to the discovery of a previously unobserved...
What if, instead of turning up the thermostat, you could warm up with high-tech, flexible patches sewn into your clothes - while significantly reducing your...
A widely used diabetes medication combined with an antihypertensive drug specifically inhibits tumor growth – this was discovered by researchers from the University of Basel’s Biozentrum two years ago. In a follow-up study, recently published in “Cell Reports”, the scientists report that this drug cocktail induces cancer cell death by switching off their energy supply.
The widely used anti-diabetes drug metformin not only reduces blood sugar but also has an anti-cancer effect. However, the metformin dose commonly used in the...
A research team from the University of Zurich has developed a new drone that can retract its propeller arms in flight and make itself small to fit through narrow gaps and holes. This is particularly useful when searching for victims of natural disasters.
Inspecting a damaged building after an earthquake or during a fire is exactly the kind of job that human rescuers would like drones to do for them. A flying...
Over the last decade, there has been much excitement about the discovery, recognised by the Nobel Prize in Physics only two years ago, that there are two types...
12.12.2018 | Event News
10.12.2018 | Event News
06.12.2018 | Event News
14.12.2018 | Power and Electrical Engineering
14.12.2018 | Physics and Astronomy
14.12.2018 | Physics and Astronomy