Psych 30 Lecture Notes: Visual Motion Perception

Looks like moving figure, but it never gets anywhere. Illusion of motion based on a theory of direction selective responses of V1 neurons. In each local region of the visual field, pattern of light changes over time to cause direction selective neurons to respond, so the whole figure appears to move.

After viewing continuous motion in the same direction for a long time, look at stationary object, and it appears to move. Also called "waterfall illusion" - look at waterfall for a while, then tree next to it appears to go up. Adaptation by motion sensitive neurons, analogous to color aftereffect.

Role of motion perception: Motion perception serves lots of helpful functions.

• Simply detecting that something is moving, draws your attention to it.
• Segmentation, foreground/background.
Motion segmentation demo Two superimposed overhead transparencies with little dots on them. Moving one allows you to see the outline shape. When stationary, can't see it.
• Provides information about 3d depth/shape and 3d trajectory. Movies (especially compelling in IMAX theaters) can give the illusion that you are moving.
Motion parallax demo Hold up two fingers (one on each hand) at different distances, move head back and forth slowly. Things that are further away slide across the retina more slowly.
• Recognition.
Johansson video: point light walkers Collection of a small number of moving dots, easily recognized as people performing various acts. No information provided other than motion of dots.

Optical Flow: Representation of the actual physical motion in the image.

Each arrow represents the speed and direction of motion for each little patch of the visual field. Near points move fast (long arrows), far points move slowly (short arrows). In this example, the arrows point away from a single point called the focus of expansion that corresponds to where the observer is heading.

First step in motion perception is for visual system to estimate optical flow from changing pattern of light in retinal image. Then 3D motions of observer and objects can be inferred from the optical flow.

Gibson hypothesis: there's sufficient information in the visual stimulus to specify unique, unambiguous interpretation of 3D motion and depth. Recently, mathematicians have proven that this hypothesis is basically correct. Caveat: distance/speed ambiguity. Small, close object when you are moving slowly projects to identical retinal images over time as large, far object when you are moving quickly. That's why you need a speedometer in your car. You're lousy at making absolute speed and distance judgements. But very good at relative speed/direction and relative distance.

Aperture problem: Visual system can estimate optical flow field only if there is enough texture in the images. Extreme example: move in front of a blank white wall, retinal image doesn't change over time, no information about motion.

Moving striped pattern (e.g. moving sinusoidal grating or striped pattern on a barber pole) is also ambiguous. Could be moving any number of different directions, and you can be biased to see one or the other of those motions depending on shape of the surrounding aperture. The ambiguity of the motion is often referred to as the aperture problem, but it really has nothing to do with the aperture per se, rather the one-dimensional nature of the the pattern.

Stimulus made by summing two gratings is unambiguous, and you see it moving in just about the right direction.

Visual Motion in the Brain

Patient known as LM who, following a stroke, had great difficulty perceiving certain types of motion. Color vision and acuity remained normal, no difficulty recognizing faces or objects, no difficulty with stereo. But cannot see coffee flowing into a cup: appears frozen like a glacier, does not perceive the fluid rising, and spills/overflows. Feels uncomfortable in a crowded room or on a street, "People were suddenly here or there, but I have not seen them moving... When I'm looking at the car first it seems far away, but then when I want to cross the road suddenly the car is very near". Lesion extends over a substantial region of visual cortex, can not localize sharply the regions relevant to motion deficit. This makes it particularly surprising that loss of motion perception can be so cleanly dissociated from other visual abilities.

V1: neurons are direction selective, respond to oriented lines/edges only when they move in the preferred direction. Direction selectivity in V1 neurons could arise via delayed inhibition from neighboring positions (see Goldstein, p. 282). But V1 neurons are not tuned for velocity of the entire pattern. They respond to the individual motion of the oriented components of the pattern, not to the motion of the pattern as a whole. Confound stimulus motion with stimulus pattern (orientation, contrast), like the ambiguous motion of the baber pole illusion.

MT: Area MT is one of the most studied regions of the cortex of the brain, probably second only to V1. Current opinion is that optical flow field is represented by neurons in area MT. MT neurons receive inputs from direction selective neurons in V1, combine multiple orientations like intersection of constraints when superimposing two moving gratings. MT neurons are velocity selective, each responds best to motion of a preferred velocity (speed and direction) within its receptive field, pretty much independent of stimulus pattern.

Record from MT neuron while displaying dots moving in different directions. Audio track allows you to hear the spikes - each click corresponds to an action potential. Neuron is strongly direction selective, responds only when dots move in a narrow range of directions.

Columnar architecture in MT for stimulus motion, neurons with similar motion preferences nearby one another, orderly progression from one motion direction to the next as you move through MT, analogous to orientation columns in V1.

MT and motion perception: Bill Newsome (Stanford Neurobiologist) and colleagues trained monkeys to perform a difficult motion discrimination task.

Monkey views moving dots and must decided which way they went, e.g., either up or down. Then inserted microelectrodes into MT to electrically stimulate the monkeys' brains while they were performing the task. When electrode is in a column selective for upward motion, stimulation biases the monkey to report up more often. Very cool result: stimulating MT neurons directly influences behavior, presumably because it influences the monkeys' conscious percepts.

Human MT: A homologous area has been localized in the human brain.

Flickering checkerboard stimulus evokes brain activity in V1 at back of head. Moving dot stimulus again evokes brain activity in V1, but also in lateral region just behind your ears - believed to be the human homologue of monkey area MT.

MST: Monkey area MST, right next to MT.

MST neurons have very large receptive fields, respond selectively to complex optical flow fields: expansion, contraction, rotation. Perhaps involved in 3D motion perception, inferring 3D motion of objects/observer from optical flow.

Neurons in area STS respond selectively to biological motion, like Johannson's point-light walkers.

Eye Movements and Motion Perception

Answer: visual images are combined with other information to inform you about motion of your eyes, head, and body. Vestibular system provides information about motion of head and body. Copy of eye movement from eye movement centers in the brain stem provides information about eye movements. Vision combined with vestibular and eye movement signals in Area MST.

Push gently on the side of your eye and the world appears to jiggle around. Helmholtz first to do this experiment and come up with a theory of how eye movement information is combined with change in retinal image to yield motion percept. Generally speaking, brain is divided into motor areas and sensory areas. The corrolary discharge is a copy of the motor signal that is transmitted to a comparator - a hypothetical structure that receives both teh corrolary discharge and the sensory movement signal (maybe in MST). If the visual motion signal is the same as the eye movement command, then you don't "see" motion. If the visual motion signal is different from the eye movement command, then you do see motion.