Uncanny Valley

From Wikipedia, the free encyclopedia

The Uncanny Valley is a hypothesis about robotics concerning the emotional response of humans to robots and other non-human entities. It was introduced by Japanese roboticist Masahiro Mori in 1970. It states that as a robot is made more humanlike in its appearance and motion, the emotional response from a human being to the robot will become increasingly positive and empathic, until a point is reached beyond which the response quickly becomes strongly repulsive. However, as the appearance and motion continue to become less distinguishable from a human being's, the emotional response becomes positive once more and approaches human-human empathy levels.

This gap of repulsive response aroused by a robot with appearance and motion between a "barely-human" and "fully human" entity is called the Uncanny Valley. The name captures the idea that a robot which is "almost human" will seem overly "strange" to a human being and thus will fail to evoke the requisite empathetic response required for productive human-robot interaction.

1 Theoretical basis
2 The Uncanny Valley in film
3 The Uncanny Valley in art
4 The Uncanny Valley in games
5 Uncanny Valley as an analogy outside artificial life
6 References
7 See also
8 External links and references

[edit] Theoretical basis

Hypothesized emotional response of human subjects is plotted against anthropomorphism of a robot, following Mori's statements. The Uncanny Valley is the region of negative emotional response towards robots that seem "almost human". Movement amplifies the emotional response.

The phenomenon can be explained by the notion that, if an entity is sufficiently non-humanlike, then the humanlike characteristics will tend to stand out and be noticed easily, generating empathy. On the other hand, if the entity is "almost human", then the non-human characteristics will be the ones that stand out, leading to a feeling of "strangeness" in the human viewer.

Another possibility is that affected individuals and corpses exhibit many visual anomalies similar to the ones seen with humanoid robots and so elicit the same alarm and revulsion. The reaction may become worse with robots since there is no overt reason for it to occur, whereas distaste for the sight of a corpse is an easy feeling to understand. Behavioural anomalies are also indicative of illness, neurological conditions or mental dysfunction and again evoke acutely negative emotions.

It can also be explained in terms of evolutionary psychology. Entities that fall in the uncanny valley are humanlike enough to be seen as part of the human being species. According to evolutionary psychology, throughout millions of years, natural selection would have logically favored features in the brain that provide a high capacity to sense and be repulsed by macro and micro-anomalies in the overall appearance of a member of the same species that reveal genetic disorders or a lack of genetic fitness. So unconsciously (and also consciously if one explicitly thinks about it), we might be alarmed by the potential impact that these abnormal humanlike entities could have on the human beings gene pool. This could explain why it is particularly disturbing for the human eye to see these humanlike entities engaging in sexual activity (see below).

Some roboticists have heavily criticized the theory, arguing that Mori had no basis for the rightmost part of his chart, as human-like robots are only now technically possible (and still only partially). David Hanson, a roboticist who developed a realistic robotic copy of his girlfriend's head, said the idea of the Uncanny Valley was "really pseudoscientific, but people treat it like it is science." Sara Kiesler, a human-robot interaction researcher at Carnegie Mellon University, questioned Uncanny Valley's scientific status, noting that "we have evidence that it’s true, and evidence that it’s not."

[edit] The Uncanny Valley in film

Although originally resulting from experimental data and applied only to robotics, the principle has been extensively applied to computer animation characters. American film critic Roger Ebert has applied the notion of Uncanny Valley to the use of make-up and costumes of humanoid creatures in movies and the deliberate minimalistic depictions of people in works of anime, specifically in Grave of the Fireflies.

The Uncanny Valley was considered by some to be the reason behind the difficulty in creating computer-animated characters. Critics of computer animated films sometimes invoke the Uncanny Valley when explaining their dislike for a particular film. The principle leads to the conclusion that to generate a positive emotional response in human beings, it is often better to include fewer human characteristics in the entity, lest it fall into the Uncanny Valley.

One example of the existence of the Valley in films is the early Pixar production Tin Toy. There, the baby shown is fully computer generated yet looks less than human and can prove frightening or unpleasant to children. The effect is lessened by the two dimensional nature of the character, but the overly defined wrinkles and (comparatively) primitive rendering of the spittle makes the character appear evil or otherworldy. A similar effect is seen in the doll's head character in the film Toy Story. The character can be again frightening to children because what it is, in essence, is an ambulatory deformed human head. Even if it is not frightening, most children prefer the cute appeal of the aliens or indeed Woody because there is less human resemblance.

In 2001, Square Pictures' photorealistic movie, Final Fantasy: The Spirits Within failed at the box-office and is often cited as a possible victim of the Uncanny Valley. The movie was the first major wide-released CGI film to feature photorealistic characters, and in turn brought about quite a bit of attention from movie critics and filmmakers alike. The Uncanny Valley theory is thought to be most prominent in Final Fantasy's character movements.

It has been said the best way to accomplish convincing human movements and to "jump" the Uncanny Valley in computer animation is to combine both motion capture and keyframing techniques. Though the former has become a popular technique, keyframing is still widely used throughout the animation industry. The film adaptation of J.R.R. Tolkien's The Lord of the Rings, which features the character Gollum, uses this combined technique with stunning results. Note, however, that Gollum's eyes and face were animated using only keyframing, and that the Gollum animation also featured other advances in modelling (including skin texture, and effects such as saliva around lips) which allow the character's external appearance to reach the far side of the Valley. Nevertheless, one obvious caveat with Gollum is that the character is evidently non-human (and, indeed, intentionally uncanny from the start) and so may not trigger the same response as a human figure would when modelled using the same techniques.

Despite advances in computer animation, some feel the Uncanny Valley affected two CGI films of 2004, The Incredibles and The Polar Express. The close dates of release led to many critics' comparison of the two movies, with some preferring the deliberately stylized appearance of the characters in The Incredibles over the more human-like characters in The Polar Express (which were described by many critics as being "disturbing"). Another reason believed to contribute to the unnerving look of the characters in 'the Polar Express' was the use of Performance Capture technology. This technology, when applied to 3D characters has been seen as having a large part in the failure of such films as 'the Polar Express' citing a disconnect between the audience and the characters. This has not been the case in studios which still employ traditional techniques to 3D models, such as Pixar in 'the Incredibles', which some believe allows them to avoid entering the Uncanny Valley.

The CG animation Final Flight of the Osiris in the Animatrix also suffers from this - although, as this was also made with the same techniques as in Final Fantasy: The Spirits Within, it was in some ways expected to. Erotic sequences with Uncanny Valley characters, as featured in Final Flight, are particularly disturbing, since they provide conflicting messages of "this is arousing" and "this is non-human".

Incidentally, Total Recall gives a good example of how uncomfortable one can feel towards Uncanny Valley encounters when Arnold Schwarzenegger attempts to pass through martian customs disguised as a fat woman.

The Uncanny Valley is also a plot point in some movies about robotics. A.I.: Artificial Intelligence centers on a future where many people are disturbed at how realistic the new line of androids is. For example, a rowdy crowd that rejoices in watching robot destruction derbies called "Flesh Fairs" falls silent when the next subject about to be ripped apart appears to be an adorable human boy. In I, Robot, the newest wave of U.S. Robotics robots is far more humanoid in facial expressions and appearance. This disturbs main character Det. Del Spooner, who was already bothered by the boxy metal robots that preceded them. "Why do you give them faces?" he asks one of the robots' programmers as he stares into a sea of identical new robots. He then discharges his firearm into the "face" of a robot at point-blank range, effectively making his point to the movie audience who will gasp at the sight of him "executing" a "person."

The Rotoscoped film, A Scanner Darkly uses the Uncanny Valley to serve its drugged up atmosphere, making the visuals at once beautiful and disturbing.

[edit] The Uncanny Valley in art

Author Scott McCloud offered a somewhat different version in his dissection of comic art, Understanding Comics. McCloud demonstrated that simpler cartoon figures leave more room for interpretation and elicit greater empathy from the audience. A cartoon's simplicity and iconic form is closer to the way we understand our own internal shape. As a drawn figure takes on more detail and approaches photorealism, the more it becomes a representation of an 'other' and so becomes more difficult for the viewer to identify with.

It is important to note that McCloud is dealing primarily with representational art, such as video games, comics and animation, and not encountering simplified humanoid entities in the 'real' world as Masahiro Mori was examining. A short essay by Mori, however, does concern the subject of art: he suggests that artistic idealizations of the human figure may actually be higher up the right-hand curve of his graph than an actual human face (citing traditional representations of the Buddha as an example).

[edit] The Uncanny Valley in games

A virtual Tiger Woods running in real-time on a PlayStation 3

As videogame consoles and personal computers get more powerful, gamers will be seeing more of the Uncanny Valley, especially on "next-generation" hardware.

Graphic chip manufacturer Nvidia and ATI (now AMD) routinely creates tech demos that show ever more realistic human characters^[1].

At past E3s, Sony has demonstrated tech demos running on upcoming hardware that creeps into the Uncanny Valley. In 2000, Sony demonstrated the Playstation 2 emotion engine with a realistic old man's face. In 2005, Sony promoted their Playstation 3 with a demo of actor Alfred Molina's face with realtime lighting, subsurface scattering, and realistic expressions. In 2006, Electronics Arts showed a computer generated head of Tiger Woods using a new type of motion capture technique which they are marketing as Playable Universal Capture (or UCAP for short ^[2])

Mary Smith from Heavy Rain, which is a game for the PlayStation 3, Xbox 360 and PC, running in realtime

Another demo was from the upcoming working title Heavy Rain. The movie shows a woman auditioning for a part in the game.^[3]

[edit] Uncanny Valley as an analogy outside artificial life

In ESPN's "Page 2" feature ^[4], columnist Patrick Hruby explained Uncanny Valley in the traditional sense, noting that in the video game Madden NFL 06 the players exhibit the unnervingly almost-human features that plague many CGI films. The column extended the Uncanny Valley term to the well-documented analogous debate over what sort of sports fan suffers more, one who roots for a perennial loser or one who roots for a perennial second-best. Hruby suggests that the perennial second-best fans, like the Red Sox Nation, suffer more because of the "Uncanny Valley" between the team's clear potential for perfection and their history of falling just short of the championship.

An article on drunkenblog ^[5] uses the Uncanny Valley analogy to describe the frustration many computer programmers experience when using the AppleScript programming language.

Transsexuals who are in-transition between genders often experience an Uncanny Valley as their gender presentation shifts from their birth sex to their identified gender. As the transitioner develops their new gender presentation, strangers' reactions exhibit the Uncanny Valley behavior. The low point occurs when the transitioner presents a disquieting mix of attributes from both genders, as happens when he or she has not yet corrected or mastered all of the major aspects of the new gender presentation. At that point the early condescending tolerance gives way to disquiet, discomfort, even hatred. With further effort, practice, and sometimes surgery, the transitioner eventually emerges triumphant on the far edge of the curve.

Some genderqueer people revel in the aforementioned Uncanny Valley of gender presentation, finding fulfillment by provoking others into questions, introspection, or alarm.

Another example is the lifestyle of people living in a foreign country. In the adaptation process to a new culture, especially if the culture is significantly different from the one a person is used to, native people's reactions to a foreigner are somewhat mimicked by the uncanny valley. At first, when the foreign person acts significantly differently enough from native people, the foreigner will be praised for trying to fit in (e.g., he or she will be at the top of the first curve), and when the foreigner has adapted to the native culture completely, he or she will fit in. But, before that time, there is an awkward area in which the native people expect the foreigner to act like them, but the foreigner is not yet completely able to do that: the uncanny valley. As an example, the transition from Western European culture to the culture of the United States might put a European in the middle of the uncanny valley, whereas if he or she had experienced an Asian culture, he or she would be instead at a point in the first curve, before the uncanny valley.