it presents you
1.Introduction to Artificial Intelligence
2.History and Evolution
3.Speech synthesis
4.Robots and Image processing
5.Sensor fusion
6.Innovation in Artificial Intelligence
7.conclusion
2. AGENDA
⢠Introduction to Artificial
Intelligence
⢠History and Evolution
⢠Speech synthesis
⢠Robots and Image processing
⢠Sensor fusion
⢠Innovation in Artificial Intelligence
⢠conclusion
3. ARTIFICIAL INTELLIGENCE
⢠To think like human
⢠To act like human
⢠To think rationally(logically)
⢠To act rationally.
4. ARTIFICIAL INTELLIGENCE
METHODS
⢠SYMBOLIC AI:
Focus on development
of knowledge based system.
⢠COMPUTATIONAL INTELLIGENCE:
Neutral networks,
fuzzy systems and evolutionary
computing.
7. HISTORY OF AI
JOHN MCCARTHY
⢠Precursors
⢠The birth of AI(1952-1956)
⢠The golden years(1956-1974)
⢠The first AI winter(1974-1980)
⢠Boom(1980-1987)
⢠Bust: second AI winter(1987-1993)
⢠AI(1993-2001)
⢠Deep learning. big data & artificial general
intelligence(2001-current).
8. GOALS OF AI
⢠Reasoning, problem solving
⢠Knowledge representation
⢠Planning
⢠Learning
⢠Natural language processing
9. TOOLS OF AI
⢠Search and optimization
⢠Logic
⢠Probabilistic methods for
uncertain reasoning
⢠Classifiers and statistical
learning methods
⢠Control theory
⢠Evaluating progress
10. EXPLOSIVE GROWTH OF AI
⢠Growth in positive side:
Useful to society.
⢠Growth in negative sides:
Harmful to society.
11. APPLICATIONS
⢠Robotics
⢠Heavy industries
⢠Medicines
⢠Telecommunications
⢠Gaming
⢠Satellite control
⢠Military activity control
⢠Network management
13. WHAT IS SPEECH SYNTHESIS?
⢠process of converting an acoustic
signal to a set of words.
⢠serve as the input to further linguistic
processing
⢠achieve speech understanding
14. SPEECH PROCESSING
⢠Signal processing:
Convert the audio wave into a sequence of feature
vectors
⢠Speech recognition:
Decode the sequence of feature vectors into a
sequence of words
⢠Semantic interpretation:
Determine the meaning of the recognized words
⢠Dialog Management:
Correct errors and help get the task done
⢠Response Generation
What words to use to maximize user understanding
⢠Speech synthesis (Text to Speech):
Generate synthetic speech from a âmarked-upâ
word string
15. What you can do with Speech
Recognition?
⢠Transcription
dictation, information retrieval
⢠Command and control
data entry, device control, navigation, call
routing
⢠Information access
airline schedules, stock quotes, directory
assistance
⢠Problem solving
travel planning, logistics
16. Transcription and Dictation
⢠Transcription:
transforming a stream of human speech into
computer-readable form
Example:
Medical reports, court proceedings, notes
Indexing (e.g., broadcasts)
⢠Dictation:
interactive composition of text
Example:
Report, correspondence, etc.
17. speech recognition system:
⢠Pre-processing:
conversion of spoken input into a
form the recogniser can process.
⢠Recognition:
identification of what has been
said.
⢠Communication:
to send the recognised input to
the application that requested it.
18. USERS AND SPEAKER
MODELS:
Different Kinds of Users
One time vs. Frequent users
Homogeneity
Technically sophisticated
speaker models
Speaker Dependent
Speaker Independent
Speaker Adaptive
19. Speech Recognition as Assistive
Technology
⢠Main use is as alternative Hands Free Data
entry mechanism
⢠Very effective
⢠Much faster than switch access
⢠Mainstream technology
⢠Used in many applications where hands are
needed for other things e.g. mobile phone
while driving, in surgical theatres
20. Speech recognition and
understanding
⢠Sphinx system
â speaker-independent
â continuous speech
â large vocabulary
⢠ATIS system
â air travel information retrieval
â context management
21. Applications:
⢠Speech Recognition
Figure out what a person is saying.
⢠Speaker Verification
Authenticate that a person is who she/he claims to
Limited speech patterns
⢠Speaker Identification
Assigns an identity to the voice of an unknown person.
Arbitrary speech pattern
22. ABILITY TO DREAM
⢠MEMORY FORMATION
â˘SUPERVISED LEARNING & UNSUPERVISED
LEAENING
â˘SUPERVISED-TRAINING DATA
â˘UNSUPERVISED-HUMAN ACTS
26. o ASSESING PEOPLE HEALTH THROUGH SENSORS
o CHEMICAL SENSORS & SILLICON RUBBERS
o MICROFLUIDICS
o DECTECTORS CAN BE CONNECTED TO
SMARTPHONES
o EXTEND TO OTHER FLUIDS LIKE TEARS &
SALAIVA
31. KURI
⢠MAYFIELD ROBOTICS
⢠LASER DEPTH PRECISION
SYSTEM
⢠REMOTE CONTROLLED VIA
KURI APP
⢠BUILT IN CAMERA
⢠MOVES, LISTENS
⢠SPEAKS & ENTERTAINS
34. Multi-sensor Fusion and Integration
ď The synergistic combination of data from multiple
sensors
ď Provide more reliable and accurate information
ď Sensor data can be incomplete, erroneous and
uncertain
35. Types of multi-sensor data
fusion
1.Complementary Fusion:
Resolves incompleteness of sensor data.
E.g. fusion of several range sensors pointed in different directions.
2.Competitive Fusion:
Fusion of uncertain sensor data from several sources
E.g. heading from odometer and magnetic compass. Reduces the
effect of uncertain and erroneous measurements.
3.Cooperative Fusion:
E.g. a touch sensor refines the estimated curvature of an object
previously sensed by range sensors
36. Architecture for a Multi-sensor Data Fusion
System
Generic multi-sensor data fusion architecture
38. Integration with three different
types of sensory processing
1.Fusion:
Sensor registration converts the sensor data common internal
representation
2. Separate Operation:
Data provided by a sensor may be significantly Different
Influences the other sensors indirectly via the system
controller and the world model.
3.Guiding or Cueing:
data from one sensor is used to guide or cue the operation of
the other sensors e.g. tactile bump sensors, IR light sensors
4.Sensor selection:
used to select the most appropriate configuration of sensors to
suit the environment conditions
39. Estimation methods
Usage:
Signal level fusion
Non-recursive:
ďˇ Weighted Average
ďˇ Least Squares
Recursive:
ďˇ Kalman Filtering
ďˇ Extended Kalman Filtering
Classification methods
Usage:
Extracting features & matching
at pixel and feature level fusion
ďˇ Parametric Templates
o Match extracted features to classes in a
multidimensional feature space
ďˇ Cluster Analysis
o Similar to SOFM
o Learn geometrical relationships
MULTI âSENSOR FUSION ALGORTHIMS
40. Classification methods Learning Vector Quantization (LVQ)
o Another type of NN
K-means Clustering
o Competitive NN
Kohonen Feature Map (SOFM)
ART,ARTMAP,Fuzzy-ART Networks
Inference methods
Usage:
Symbol level fusion â
evidential reasoning
ďˇ Bayesian Inference
o Information combined according to the rules
of probability theory
o Bayes formula
o between sample data sets
ďˇ Dempster-Shafer Method
o Rectifies some instances where probabilities
may become unstable in Bayesian inference
ďˇ Generalized Evidence Processing
o Unifies Bayesian and Dempster- Shafer
methods
41. A
Artificial intelligence methods
Usage:
Can be used at different levels
of fusion
ďˇ Expert System
o Performs inferences using a data set and rule-
based knowledge base
ďˇ Neural Networks
o Adaptive
o Backpropagation
ďˇ Fuzzy Logic
o Multiple-valued logic where variables are
assigned degrees of membership between 0
and 1
âmay beâ exists between "yesâ and ânoâ
42. 12
Output to system controller
Symbol
Level
Feature
Level
Signal&
Pixel
Level
Implementation of target tracking system integrating visual
detection and ultrasonic sensory data