By making a neural-network computer model that can be fooled by optical illusions like humans, the researchers advanced knowledge of the human visual system and may help improve artificial vision
Is that circle green or gray? Are the center lines straight or tilted?
Optical illusions can be fun to experience and debate, but understanding how human brains perceive these different phenomena remains an active area of scientific research. For one class of optical illusions, called contextual phenomena, those perceptions are known to depend on context.
For example, the color you think a central circle is depends on the color of the surrounding ring. Sometimes the outer color makes the inner color appear more similar, such as a neighboring green ring making a blue ring appear turquoise -- but sometimes the outer color makes the inner color appear less similar, such as a pink ring making a grey circle appear greenish.
A team of Brown University computer vision experts went back to square one to understand the neural mechanisms of these contextual phenomena. Their study was published on Sept. 20 in Psychological Review.
"There's growing consensus that optical illusions are not a bug but a feature," said Thomas Serre, an associate professor of cognitive, linguistic and psychological sciences at Brown and the paper's senior author. "I think they're a feature. They may represent edge cases for our visual system, but our vision is so powerful in day-to-day life and in recognizing objects."
For the study, the team lead by Serre, who is affiliated with Brown's Carney Institute for Brain Science, started with a computational model constrained by anatomical and neurophysiological data of the visual cortex. The model aimed to capture how neighboring cortical neurons send messages to each other and adjust one another's responses when presented with complex stimuli such as contextual optical illusions.
One innovation the team included in their model was a specific pattern of hypothesized feedback connections between neurons, said Serre. These feedback connections are able to increase or decrease -- excite or inhibit -- the response of a central neuron, depending on the visual context.
These feedback connections are not present in most deep learning algorithms. Deep learning is a powerful kind of artificial intelligence that is able to learn complex patterns in data, such as recognizing images and parsing normal speech, and depends on multiple layers of artificial neural networks working together. However, most deep learning algorithms only include feedforward connections between layers, not Serre's innovative feedback connections between neurons within a layer.
Once the model was constructed, the team presented it a variety of context-dependent illusions. The researchers "tuned" the strength of the feedback excitatory or inhibitory connections so that model neurons responded in a way consistent with neurophysiology data from the primate visual cortex.
Then they tested the model on a variety of contextual illusions and again found the model perceived the illusions like humans.
In order to test if they made the model needlessly complex, they lesioned the model -- selectively removing some of the connections. When the model was missing some of the connections, the data didn't match the human perception data as accurately.
"Our model is the simplest model that is both necessary and sufficient to explain the behavior of the visual cortex in regard to contextual illusions," Serre said. "This was really textbook computational neuroscience work -- we started with a model to explain neurophysiology data and ended with predictions for human psychophysics data."
In addition to providing a unifying explanation for how humans see a class of optical illusions, Serre is building on this model with the goal of improving artificial vision.
State-of-the-art artificial vision algorithms, such as those used to tag faces or recognize stop signs, have trouble seeing context, he noted. By including horizontal connections tuned by context-dependent optical illusions, he hopes to address this weakness.
Perhaps visual deep learning programs that take context into account will be harder to fool. A certain sticker, when stuck on a stop sign can trick an artificial vision system into thinking it is a 65-mile-per-hour speed limit sign, which is dangerous, Serre said.
The research team included Brown graduate student David Mély and postdoctoral scholar Drew Linsley. The research was supported by the National Science Foundation (IIS-1252951) and DARPA (YFA N66001-14-1-4037).
Touchscreens go 3D with buttons that pulsate and vibrate under your fingertips
14.03.2019 | Universität des Saarlandes
EU project CALADAN set to reduce manufacturing cost of Terabit/s capable optical transceivers
11.03.2019 | IHP - Leibniz-Institut für innovative Mikroelektronik
Nano- and microtechnology are promising candidates not only for medical applications such as drug delivery but also for the creation of little robots or flexible integrated sensors. Scientists from the Max Planck Institute for Polymer Research (MPI-P) have created magnetic microparticles, with a newly developed method, that could pave the way for building micro-motors or guiding drugs in the human body to a target, like a tumor. The preparation of such structures as well as their remote-control can be regulated using magnetic fields and therefore can find application in an array of domains.
The magnetic properties of a material control how this material responds to the presence of a magnetic field. Iron oxide is the main component of rust but also...
Due to the special arrangement of its molecules, a new coating made of corn starch is able to repair small scratches by itself through heat: The cross-linking via ring-shaped molecules makes the material mobile, so that it compensates for the scratches and these disappear again.
Superficial micro-scratches on the car body or on other high-gloss surfaces are harmless, but annoying. Especially in the luxury segment such surfaces are...
The Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI) at the Large Binocular Telescope (LBT) in Arizona released its first image of the surface magnetic field of another star. In a paper in the European journal Astronomy & Astrophysics, the PEPSI team presents a Zeeman- Doppler-Image of the surface of the magnetically active star II Pegasi.
A special technique allows astronomers to resolve the surfaces of faraway stars. Those are otherwise only seen as point sources, even in the largest telescopes...
Researchers at Chalmers University of Technology and the University of Gothenburg, Sweden, have proposed a way to create a completely new source of radiation. Ultra-intense light pulses consist of the motion of a single wave and can be described as a tsunami of light. The strong wave can be used to study interactions between matter and light in a unique way. Their research is now published in the scientific journal Physical Review Letters.
"This source of radiation lets us look at reality through a new angle - it is like twisting a mirror and discovering something completely different," says...
New research group at the University of Jena combines theory and experiment to demonstrate for the first time certain physical processes in a quantum vacuum
For most people, a vacuum is an empty space. Quantum physics, on the other hand, assumes that even in this lowest-energy state, particles and antiparticles...
11.03.2019 | Event News
01.03.2019 | Event News
28.02.2019 | Event News
21.03.2019 | Life Sciences
21.03.2019 | Physics and Astronomy
21.03.2019 | HANNOVER MESSE