310 likes | 341 Views
Explore cutting-edge algorithms and techniques in computer animation, focusing on behavioral animation and crowd management. Learn about individual and aggregate behaviors, reactive behavior, and intelligent behavior modeling. Discover approaches like flocking, navigation, reasoning, perception, and modeling flight to create realistic animations. Dive into advanced topics such as modeling individuality, personality, emotions, and internal state models. Gain insights into emergent behavior, pedestrian traffic simulation, and managing crowds effectively for entertainment purposes.
E N D
Computer AnimationAlgorithms and Techniques Behavioral Animation Rick Parent Computer Animation
http://www.gamasutra.com/view/feature/3405/ai_for_games_and_animation_a_.phphttp://www.gamasutra.com/view/feature/3405/ai_for_games_and_animation_a_.php Rick Parent Computer Animation
Behavioral Animation Individuals Reactive behavior Intelligent behavior Personality Groups Aggregate behavior Crowd behavior Rick Parent Computer Animation
Reactive Behavior If <condition> then <action> current conditions - usually omniscence simple action - carried out directly and immediately problems of scale: conflicting rules priority selection or ordering Rick Parent Computer Animation
Aggregate Behavior Typical qualities Rick Parent Computer Animation
Primitive Behavior - Flocking Local control – for realism, the flock member only reacts to locally accessible information Perception – fov vision – angle can change with speed Interacting with other members – stay with friends, avoid bumping into each other Interacting with the environment – collision avoidance is primary Rick Parent Computer Animation
Primitive Behavior - Flocking Original work by Craig Reynolds Global control – need control of flock script flock leader global migratory urge Negotiating the motion Collision avoidance – steer to avoid Splitting and rejoining – difficult to tune parameters Modeling flight – e.g., banking into turns Rick Parent Computer Animation
Forces Or “Reasoning” (e.g. rule-based) Negotiating the Motion Rick Parent Computer Animation
Navigating Obstacles: Using repulsive forces Rick Parent Computer Animation
Navigating using bounding sphere Rick Parent Computer Animation
Navigating Testing for being on a collision path with (bounding) sphere Given: P, V, C, r Rick Parent Computer Animation
Finding closest non-colliding point Rick Parent Computer Animation
Navigating – finding a pass Render in z-buffer Sample environments with rays Rick Parent Computer Animation
Modeling Flight –common in flocking Rick Parent Computer Animation
Modeling Flight Rick Parent Computer Animation
Modeling Flight Rick Parent Computer Animation
Modeling Flight Rick Parent Computer Animation
Primitive Behavior - Prey-Preditor unbalanced abilities maximum velocity maximum acceleration maximum angular acceleration maximum angular velocity vision - distance, movement, fov Rick Parent Computer Animation
Intelligent behavior Reasoning Causation Knowledge about environment Vision - immediately sensed Memory - previous experience Rick Parent Computer Animation
Vision Omniscience – access to the database FOV vision – culling but no occlusions Occluded vision ray casting – sample environment z-buffer using object IDs as z’s Object recognition - ? Rick Parent Computer Animation
Memory What is recorded about the environment Spatial occupancy Transient events: time-stamps Rick Parent Computer Animation
Intelligent Behavior Autonomous behavior ‘Self-animated’ characters Perception & reasoning about environment Personality, emotions, dispositions Manifestations of Individuality Body Expressions and Gestures Facial expressions Speech Rick Parent Computer Animation
Internal State Models what the agent needs to do Suggested precedence classes of internal state variables Imperatives Desires Suggestions Rick Parent Computer Animation
Arbitration among competing intentions Levels of Behavior Hooks for the animator to impose control Rick Parent Computer Animation
Expressions and Gestures ToBI – Tones and Break Indices LMA – Laban Movement Analysis BEAT EMOTE RUTH Greta Rick Parent Computer Animation
Modeling Individuality Personality – long term qualities Emotions – short term Mood – third level? Basic emotions?: happy, sad, fear, disgust, surprise, anger Models of personality from psychology OCEAN: openness, conscientiousness, extroversion, agreeableness, neuroticism PEN: extraversion, neuroticism, psychoticism OCC: how perceptions dictate emotional experience Rick Parent Computer Animation
Modeling Individuality Improv AlphaWolf Rick Parent Computer Animation
Crowd Management Emergent behavior Statistical behavior v. believable individual behaviors Homogeneous activity v. Internal structure For evaluation Pedestrian traffic simulation Traffic flow Emergency response modeling For entertainment Background crowds Rick Parent Computer Animation
Emergent behavior: similar to flocking collision avoidance ‘intelligent’ paths Crowds From a distance: statistical behavior nonsensical detailed motion reasonable visual effect Internal structure limited interaction among members group formation Rick Parent Computer Animation