Uncanny Valley Effect

fMRI studies show that when subjects view humanoid figures in the uncanny valley range, both the amygdala and the prefrontal cortex in the brain are activated simultaneously.

The core idea of the uncanny valley theory is that as a simulated object increasingly resembles a real person, people's sense of affinity first rises, but when it reaches a critical point where subtle abnormalities still exist, affinity suddenly drops into a valley, producing a sense of eeriness and rejection.

The uncanny valley effect refers to the phenomenon whereby as a robot's appearance increasingly resembles a real human, human affinity first rises, then plunges sharply into discomfort or even fear at the threshold where it approaches but does not fully achieve realism, only recovering when the resemblance advances further to the point where it is indistinguishable from a real person.

When AI avatars closely resemble but are not fully equivalent to real humans, people experience discomfort and aversion (the uncanny valley effect).

Real-time rendering is more likely to trigger the uncanny valley effect than static images, because character facial details are constantly changing, and the deviation of each AI inference is continuously amplified.

The more realistic a digital human is, the more likely even minor anomalies will trigger viewer discomfort, meaning technological progress does not necessarily lead to a linear improvement in acceptance.

The human visual perception system is extremely sensitive to unnaturally 'perfect' motion, producing a rejection response even when one cannot explicitly identify the problem.

Real human motion naturally contains a large amount of small, irregular 'noise,' such as natural acceleration and deceleration during head turns, slight tremors, and imperfect motion trajectories.

When AI edits or generates motion, it tends to produce overly smooth and overly precise motion curves, and this 'perfection' causes the human brain to judge the result as 'unreal.'

AI-generated food images can trigger a visual aversion similar to the uncanny valley effect in robotics, where the closer they are to reality yet with subtle deviations, the more discomfort they tend to cause.

The uncanny valley effect also applies to AI-generated content: when automated videos approach but fall short of professional quality, they may trigger discomfort and reduce trust.

Overly templated auto-generated demos may trigger discomfort similar to the 'uncanny valley effect,' undermining a product's uniqueness.

The Uncanny Valley effect, proposed by Masahiro Mori, points out that when an artificial character approaches but fails to fully achieve a truly human appearance, viewers experience a strong sense of discomfort.

The higher the visual fidelity of a scene, the more any remaining physical logic flaws are amplified manifold, producing a strong sense of dissonance.

The uncanny valley effect was originally proposed by Japanese roboticist Masahiro Mori in 1970 to describe human emotional responses to humanoid robots.

The uncanny valley effect was proposed by Japanese roboticist Masahiro Mori in 1970, referring to the phenomenon in which human affinity for robots suddenly drops into negative territory as robots approach but do not fully reach human likeness.

现实主义剧情片是AI影视最难攻克的领域，因对恐怖谷效应最为敏感

AI生成的真人风格视频往往比风格化动画更容易引发观众排斥反应