You are on page 1of 34

Imitation of Natural world in 3D Gaming

History and Benefits

Machine Learning
Rule based approach is used in research areas. Very costly and specific devices and instruments are used. Try to develop the enginering devices that can replace or can be used more efficiently than the human.

Machine Learning
Observations are done in very constrained environment. Then results are applied in real world environments for better results. If the constraints of real world doesnt matches the ideal situations then the research is set as rule. But what of real world problems of society.

A number of talented researches are going on parallel with developments of engineering works in various fields. We would just try to suggest some point if two researches of separate areas can go together.

History of Gaming
1947:

Cathode Ray Tube Amusement Device The earliest known interactive electronic game was by Thomas T. Goldsmith Jr. and Estle Ray Mann on a cathode ray tube.[3] The patent was filed on January 25, 1947 and issued on December 14, 1948. 1950-1951: Chess In March 1950, Claude Shannon devised a chess.

1951: NIM 1952: OXO / Noughts and Crosses (Tic-Tac-Toe) 1958: Tennis for Two 1961: Spacewar! 1966: Odyssey 1971: Galaxy Game 1971: Computer Space 1972: Pong

A drawing of the NIMROD computer.

Tennis for Two recreation

Classifiction of Gaming

Classification
Core

Games Casual Games Serious Games

Classifiction of Gaming

On the basis of

Platform

PC Games Console Games Handheld Games Arcade Games method of game play, types of goals, art style and more. fostered more life-like and complex games which have in turn introduced or enhanced genre possibilities (eg, virtual pets), pushed the boundaries of existing video gaming or in some cases add new possibilities in play (such as that seen with titles specifically designed for devices like Sony's EyeToy or Nintendo's Wii Remote. massively multiplayer online role-playing games, or, more commonly, MMORPGs.

Genre

3D Gaming

The 1990s were a decade of marked innovation in video gaming. It was a decade of transition from pixels to full fledged 3D graphics and gave rise to several genres of video games including, but not limited to, the first person shooter, real time strategy, survival horror, and MMO. Handheld gaming began to become more popular throughout the decade, thanks in part to the release of the Game Boy. Arcade games, although still relatively popular in the early 1990s, begin a decline as home consoles become more common.

Artificial Intelligence

What is artificial intelligence? It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable. Yes, but what is intelligence? Intelligence is the computational part of the ability to achieve goals in the world. Varying kinds and degrees of intelligence occur in people, many animals and some machines. Isn't there a solid definition of intelligence that doesn't depend on relating it to human intelligence? Not yet. The problem is that we cannot yet characterize in general what kinds of computational procedures we want to call intelligent. We understand some of the mechanisms of intelligence and not others.

History of AI

1943: early beginnings

McCulloch & Pitts: Boolean circuit model of brain

1950: Turing

Turing's "Computing Machinery and Intelligence

1956: birth of AI

Dartmouth meeting: "Artificial Intelligence name adopted

1950s: initial promise


Early AI programs, including Samuel's checkers program Newell & Simon's Logic Theorist

1955-65: great enthusiasm


Newell and Simon: GPS, general problem solver Gelertner: Geometry Theorem Prover McCarthy: invention of LISP

196673: Reality dawns


Realization that many AI problems are intractable Limitations of existing neural network methods identified

Neural network research almost disappears

196985: Adding domain knowledge


Development of knowledge-based systems Success of rule-based expert systems,


E.g., DENDRAL, MYCIN But were brittle and did not scale well in practice

1986-- Rise of machine learning


Neural networks return to popularity Major advances in machine learning algorithms and applications

1990-- Role of uncertainty

Bayesian networks as a knowledge representation framework

1995-- AI as Science

Integration of learning, reasoning, knowledge representation AI methods used in vision, language, data mining, etc

Use of AI in Gaming
Finite State Machine (FSM), the Moore and Mealy machines of the 1950s scripting, or scripted behaviours. path-planning A* algorithm is heavily employed in commercial games to traverse these environment. Colin McRae Rally 2 used a AI approach to learn the rails along which the computercontrolled opponent vehicles should travel. Black & White used a BDI (Belief-Desire-Intent) model to define the NPCs.

Machine Learning
Recent years have, however, also witnessed a number of contributions which apply machine learning techniques to the problem of game AI The one-on-one martial arts game Tao Feng, by pitting an online reinforcementlearning system against the games in-built AI.

Problem.
These approaches still, however, relied upon rulebases of manually constructed state-action pairings. Solution imitation learning - that is, the automated acquisition of skills or behaviours via observation of an experts demonstration may be possible.

Imitation Research Platforms

Benefits.. Robotics and Computer Vision

Expensive motion-capture equipment


v/s

Recording of entire sessions (in the form of network traffic) to comprehensive data files Engineering challenge of achieving realistic humanoid physical motion. Recordings of gameplay sessions encode the actual behaviours of human players under rapidly-changing conditions and against opponents of comparative skill Games characters resemble idealised robot navigation worlds.

Suggestion.

Employing games as a platform for imitation-learning research gives access to almost unlimited training samples, containing no noise or other artifacts, and which encode human planning and decision-making behaviors of potentially far greater interest than the limb-motion data typically found in robotic imitation.

Imitation, Behaviour and AI


Imitation learning, as the name suggests, refers to an agents acquisition of skills or behaviours by examining a demonstrators execution of a given task. Human Imitation by Machine in AI.

Biological Imitation

Imitation and Mimicry in Infants and Animal

The fields of ethnology, neuroscience, psychology and linguistics suggest that imitation is an important component of higher intelligence.

Infants, are capable only of copying actions within their immediate perceptual field and of the same sensory modality (for instance, hand gestures where the child's own hand is also visible). The study of imitation as a form of intelligence became somewhat stigmatised.

Imitation in Animals

Simple copying, as seen in species such as rats and monkeys, refers to actions which are learned associatively and reproduced when the animal is placed in the same social context as that wherein the behavior was originally observed. More complex mimicry is observed in species such as apes and dolphins - these animals are capable of mastering sequences of manipulations, distinguishable from straightforward copying by the animals' ability to reproduce isolated subsections of (or variations upon) the imitated behaviours in contexts other than those in which they were learned.

Games-Based Imitation Learning

Some scattered examples of imitation learning in games.


Tron which allows humans to play online against computercontrolled opponents. learning simple actions from human demonstration in the FPS (first-person shooter) game Soldier of Fortune 2. He demonstrates that an MLP performs significantly better than the other approaches, at the cost of an increased learning time. There is means of automatically recognizing sequences of team behaviours using the game Unreal Tournament. Elsewhere, Zanetti and el Rhalibi employ an evolutionary neural network approach to imitation learning in the game Quake 3.

Benefits of Mixing AI Gaming imitation with other science.

Deep Blue defeated the reigning world chess champion Garry Kasparov in 1997 AI program proved a mathematical conjecture (Robbins conjecture) unsolved for decades

During the 1991 Gulf War, US forces deployed an AI logistics planning and scheduling program that involved up to 50,000 vehicles, cargo, and people
NASA's on-board autonomous planning program controlled the scheduling of operations for a spacecraft

Proverb solves crossword puzzles better than most humans


Robot driving: DARPA grand challenge 2003-2007 2006: face recognition software available in consumer cameras

Grand Challenge

Cash prizes ($1 to $2 million) offered to first robots to complete a long course completely unassisted Stimulates research in vision, robotics, planning, machine learning, reasoning, etc

2004 Grand Challenge:


150 mile route in Nevada desert Furthest any robot went was about 7 miles but hardest terrain was at the beginning of the course

2005 Grand Challenge:


132 mile race Narrow tunnels, winding mountain passes, etc Stanford 1st, CMU 2nd, both finished in about 6 hours

2007 Urban Grand Challenge

This November in Victorville, California

Stanford Racing Team

Stanley Robot

www.stanfordracing.org

SENSOR INTERFACE
RDDF database
corridor

PERCEPTION

PLANNING&CONTROL
Top level control
pause/disable command

USER INTERFACE
Touch screen UI Wireless E-Stop

Laser 1 interface
RDDF corridor (smoothed and original) driving mode

Laser 2 interface Laser 3 interface Laser 4 interface Laser 5 interface Camera interface Radar interface Road finder
laser map map trajectory road center

Path planner

Laser mapper Vision mapper Radar mapper


vehicle state (pose, velocity)

vision map

VEHICLE INTERFACE

obstacle list

Steering control Touareg interface

GPS position GPS compass

UKF Pose estimation


vehicle state (pose, velocity)

vehicle state

Throttle/brake control Power server interface

IMU interface
Wheel velocity Brake/steering
heart beats

Surface assessment

velocity limit

Linux processes start/stop health status

emergency stop

Process controller
data

Health monitor

power on/off

GLOBAL SERVICES

Data logger
Communication requests Communication channels

File system
clocks

Inter-process communication (IPC) server

Time server

Planning = Rolling out Trajectories

2004: Barstow, CA, to Primm, NV

150 mile off-road robot race across the Mojave desert Natural and manmade hazards No driver, no remote control No dynamic passing Fastest vehicle wins the race (and 2 million dollar prize)

2005 Semi-Finalists: 43 Teams

The Grand Challenge Race

HAL: from the movie 2001

2001: A Space Odyssey


classic HAL

science fiction movie from 1969

part

of the story centers around an intelligent computer called HAL HAL is the brains of an intelligent spaceship in the movie, HAL can

speak easily with the crew see and understand the emotions of the crew navigate the ship automatically diagnose on-board problems make life-and-death decisions display emotions In 1969 this was science fiction: is it still science fiction?

Days will come and we will definitely jump to get to the more complex technologies with more ease by applying games playing technologies.

Thanks!....