חדר איימס
English
Share
Overview
מקור: המעבדה של ג’ונתן פלומבאום – אוניברסיטת ג’ונס הופקינס
האתגר הקשה ביותר של תפיסה חזותית מתואר לעתים קרובות כאחד של שחזור מידע על מרחב תלת מימדי ברשתיות דו ממדיות. הרשתית היא הרקמה הרגישה לאור בתוך העין האנושית. האור משתקף מחפצים בעולם, מטילים תחזיות על הרשתית הממריצות תאים רגישים לאור אלה. חפצים זה לצד זה בעולם ייצרו גירויים זה לצד זה ברשתית. אבל חפצים מרוחקים יותר מן הצופה לא יכול לייצר גירויים רחוקים יותר, לעומת אובייקטים סמוכים כלומר. המרחק- המימד השלישי – מכווץ על הרשתית.
אז איך אנחנו רואים בשלושה ממדים? התשובה היא שהמוח האנושי מיישם מגוון הנחות והוריסטיקה כדי להסיק מסקנות לגבי מרחקים בהתחשב בתשומות שהתקבלו ברשתית. בחקר התפיסה, יש מסורת ארוכה של שימוש באשליות חזותיות כדרך לזהות כמה היוריסטיקה וההנחות האלה. אם חוקרים יודעים באילו טריקים המוח משתמש, הם אמורים להיות מסוגלים לגרום למוח לראות דברים בצורה לא מדויקת. וידאו זה יראה לך כיצד לבנות חדר איימס, אשליה חזותית הממחישה את אחת ההנחות המיושמות על ידי מערכת הראייה האנושית על מנת לשחזר עומק חזותי.
Procedure
Results
What do you see when you look into the Ames Room? Figure 3 schematizes the effect-the figurine on the right should look much larger than the one on the left, even though you know they are the same size.
Figure 3: Schematic representation of what people see in the Ames Room, compared to the fact of the matter, i.e., what is actually there. The right side of the figure shows the true relative sizes and distances of the figurines in the room: they are equally tall, and the one on the left is farther away from the viewer. But when looking through the aperture, the illusion, depicted on the left, is that the figurine on the left looks to be standing next to the person on the right, and that figurine also look much smaller.
What's going on? Before explaining the Ames Room in particular, we need to consider the general problem of perceiving size and depth at the same time. The projection that an object produces on the retina will vary in size in proportion to the size of the object; but projections will also vary in size as a function of an object's distance from the surface it projects on (in this case the retina). In other words, a very large projection on the retina could mean that the relevant object is very large and reasonably far from the observer, or it could mean that the relevant object is small, but very close by. Big objects can cast small projections when they are far away, and small objects can cast large projections when they are nearby. Separating size and distance is one of the major challenges to 3D vision. Figure 4 schematizes this general problem in reference to two trees of the same size but at different relative distances from an observer.
Figure 4: A schematic diagram to illustrate the problem of simultaneously perceiving size and distance. On the left side of the figure are two trees of equal height. To their right, is an observer, denoted as an eye. Because of the relative position of the observer to the trees, and the physics of optical projection, the tree to the left of the observer will cast a much smaller reflection on the retina of the observer, in comparison to the tree on the right. This is because the tree on the right is closer to the observer. But given the projections on the two dimensional retina, what should the observer think, is the tree on the left smaller, or just farther away?
Returning to the Ames Room now, what that specific illusions demonstrates is one of the tricks the brain uses to estimate object sizes and object distances. Specifically, the brain applies an assumption: it assumes that in the absence of strong countervailing evidence, that structures connect to one another at right-angles.
Let's look back at the Ames Room configuration now to understand how the illusion exploits that assumption. The walls of the room are all the same color, and looking through a small aperture people cannot tell that the far wall is diagonal. So the brain assumes that it is straight-that the Ames Room is rectangular as opposed to irregular. The implication of that assumption is that the figurine on the left side of the room (relative to the aperture view) is much closer to the observer than it actually is. In fact, the implication is that it is at the same distance as the figurine on the right. Because it is actually farther away, the figurine on the left projects a smaller image on the observer's retina than the figurine on the right. But the brain has made an assumption that implies that they are the same distance. So what could account for the differences in the sizes of the figurine projections? The brain get's tricked: it reasons that the figurine on the left must actually be much smaller than the one the on right producing the illusion. Figure 5 walks through this reasoning in relation to the geometry of the Ames Room.
Figure 5. The Ames Room tricks the human brain, producing the size illusion, by taking advantage of an assumption that the brain makes about geometry. Specifically, the human brain assumes that walls attach to one another at right angles. Looking through the aperture of the Ames Room, the brain cannot gather countervailing evidence, and so it applies that assumption. The result is that it thinks the Ames Room is rectangular, with the far wall occupying the positions of the dotted line in the figure. The implication, then, is that the two figurines are side-by-side, and that the left figurine is much closer to the observer than is accurate. The brain then asks itself why two side-by-side objects cast so differently sized projections on the retina. The answer it gives: they must be different sizes.
Applications and Summary
Understanding how humans perceive the visual world in 3D has been a major area of research focus and a major achievement of the modern study of perception. Some of the important applications that have arisen, as a result, are in the development of 3D and virtual reality viewing technology. The Ames Room specifically has been used for a long time in the movies, as a sort of special effect. Suppose a movie needs to depict a giant, or someone very small. Shooting interior scenes inside an Ames Room can produce the illusion for the viewer that some people are much larger (or smaller) than they actually are. The camera, is after all an aperture.
Transcript
Visual illusions often exploit the fact that a person receives input from both eyes to see the world in three dimensions.
For example, if someone is talking to a friend through their open apartment door, they’ll discern this individual as standing a few steps in front of the hallway wall. This perception is due, in part, to the fact that the person’s brain receives and combines two images of the friend—one from the right eye, and another from the left.
However, if the door suddenly slams shut and the renter is forced to look at their friend with one eye through a hole—size and depth become more difficult to judge.
The Ames Room—a structure named after its ophthalmologist inventor Aldebert Ames—manipulates this binocular aspect of vision, and cleverly uses geometry to trick the visual system into identifying objects’ sizes as being different when in fact they are the same.
This video demonstrates how to design an Ames Room, documents the illusion it creates, and discusses how such deception is used in special effects for movies.
In this experiment, participants view a distorted exhibit with a single eye by peering through an aperture—the entire setup is the Ames room. The forced perspective ensures that the brain receives only one image of the scene, and no depth cues from the contralateral eye.
The geometrical trick here is that the display is constructed with sidewalls of irregular length, making the back wall diagonal rather than parallel to the front wall. Such construction does not produce the right angles that are expected in a room.
Instead, a situation is created where two equally sized figurines—placed in the back corners—vary in distance from the aperture. That is, the one on the left is farther away than the one on the right side.
Participants are asked to report what they see, and their responses are recorded as the dependent variable. It’s expected that they will report the character on the right as being much larger, even though they are the same size.
Before participants arrive, assemble a small-scaled version of the Ames Room, using scissors or a knife, tape, a ruler, four, one-foot wide cardboard pieces of varying lengths, and two, equally sized action figures.
First make the aperture: Trace around a penny in the center of one of the foot-long pieces. Cut around this outline and then push the part out to leave a small hole. Stand this segment up, and with tape, attach it to the other one-foot long portion of cardboard to form a 90° angle on what will be its right side.
Now fasten the two-foot piece on the left side, forming another 90° angle. Finally, connect the remaining one-and-a-half-foot section to close the polygon. Note that the backside should not form right angles.
With the display assembled, place two figures of the same height in the back corners—one at each vertex.
To begin the experiment, make sure that the back is covered to avoid revealing the exhibit’s skewed geometry and seat a participant in front. Explain that she is to look into the aperture and report which of the two characters is larger.
At this point, allow the participant to lean in. While she is peering inside, ask her which one is larger and note her response. [Participant says: Right is larger.]
In this demonstration, the participant was tricked into thinking that the characters differed in size. In this case, she reported the right figure as being larger than the left one.
Here, the underlying expectation was based on previous experiences—that walls should attach at right angles. Because the brain could not gather countervailing evidence—like that which could be provided by the other eye—it therefore applied the assumption. This resulted in the room appearing rectangular, with the far wall occupying the position shown.
The implication then was that the two figures were side-by-side. Now, what appeared on the retina—two differently sized projections—was interpreted by the brain as a size difference, rather than one caused by the objects standing at different distances from the viewer.
Now that you are familiar with the Ames Room illusion, let’s look at how its principles are applied in situations involving special effects and virtual-reality viewing.
Many movie scenes have been filmed inside an Ames Room set. Because the frames are captured through the apertures of cameras, directors can depict an actor as being super small and a robot as being much larger than it really is by simply positioning them appropriately around the room.
By helping us understand the cues needed by humans to perceive the visual world in 3D, the Ames Room has also helped researchers design better virtual reality tools that employ size illusions.
Such systems can actually help individuals visualize home remodeling and design projects. By using virtual reality devices, objects can be positioned so that viewers can perceive the actuality of appearance within a given space.
You’ve just watched JoVE’s video on the Ames Room. Now you should have a good understanding of how to design the exhibit and run the demonstration, as well as how to interpret responses and apply the principles behind the illusion.
Thanks for watching!