Neural Information Systems

chickenchairwomanΤεχνίτη Νοημοσύνη και Ρομποτική

19 Οκτ 2013 (πριν από 3 χρόνια και 7 μήνες)

75 εμφανίσεις

Neural Information Systems


FACEFLOW: Face Recognition System

ANSER :Rainfall Estimating System

THONN:Date Simulation System


Dr. Ming Zhang, Associate Professor


Department of Physics, Computer Science & Engineering



Dr. Ming Zhang





ANSER System Interface






PT
-
HONN Data Simulator


FACEFLOW (1992
-

2002)



A computer vision system for recognition of


3
-
dimensional moving faces using GAT model

(neural network
G
roup
-
based
A
daptive tolerance
T
ree)



A$850,000 supported by SITA (Society Internationale
de Telecommunications Aeronautiques)


A$40,500 supported by Australia Research Council


A$78,000 supported by Australia Department of
Education.


US$160,000 supported by USA National Research
Council.





Why Develop FACEFLOW ?




-
To use new generation computer technique,
artificial neural network, for developing information
systems.

-
No real world face recognition system is running in
the world.

-
Big security market

-
Biometric system

-
ID card identification system

-
Car and house security system




What Approved

Artificial Neural Network Techniques can :


-
Can recognition one face in the laboratory
using less than 1 second


-
Currently can recognition about 1000 faces


Next Step


Rebuild

interface

for

face

recognition

system


Face

Detection


Lighting


Background


Make

up


New

neural

network

models



More complicated pattern recognition



Build a rear world face recognition System

Microsoft Visual C++. Net

Enterprise Version!

PixelSmart Image Capture Card

Source Codes
-

Compiled & Linked!

Victor Image Processing Library

Running in Visual C++.NET
!





Faceflow: Face Model Simulator

Test Different Models!

BrainMaker Neural Network Software


the Fastest Training Package!




ExploreNet Neural Network Software

The Best Interface Package!

FERET Facial Image Database

Standard Face Database!
































Research Lab

In Modern Building !

We have a pattern recognition
lab in the ARC building

We have our own room to do
research.


Dr. Ming Zhang


Neuron Network Group Models


GAT Tree Model


-

real time and real world face recognition


Neuron
-
Adaptive Neural Network Models



-

best match real world data


Center Of Motion Model
-

motion center


Second Order Vision Model
-

motion direction


NAAT Tree Model
-

a possible more powerful model
for face recognition




Research Topics

Dr. Ming Zhang


11/1999


07/2000:


Senior USA NRC Research Associate


NOAA,

Funding $70,000.


03/1995


11/1999:
Ph.D. Supervisor


University of Western Sydney


Funding: A$203,724 Cash from Fujitsu, ARC, & UWSM



07/1994
-
03/1995:
Ph.D. Supervisor

and
Lecturer



Monash University, A$50,000 Grant from Fujitsu)


11/1992
-
07/1994:

Project Manager

&
P.H.D. Supervisor


University of Wollongong, (A$850,000 from SITA)


07/1991
-
10/1992: USA NRC
Postdoctoral Fellow



NOAA, Funding: US$100,000)



07/1989
-
06/1991:
Associate Professor

and
Postdoctoral Fellow

The Chinese Academy of the Sciences. Funding: RMB$2,000,000



Dr. Ming Zhang

Dr. Ming Zhang’ s Publications


(Face Recognition)

1

Journal

Papers

1
)


Ming

Zhang,

Rex

Gantenbein,

Sung

Y
.

Shin,

and

Chih
-
Cheng

Hung,

The

application

of


artificial

neural

networks

in

knowledge
-
based

information

systems,

International

Journal


of

Computer

and

Information

Science,

Vol

2
,

No
.
2
,

2001
,

pp
.
49

-

58
.


2
)


Ming

Zhang,

Jing

Chung

Zhang,

John

Fulcher,

"Neural

network

group

models

for

data

approximation",

International

Journal

of

Neural

Systems,

Vol
.

10
,

No
.

2
,

April,

2000
,

pp
.

123
-
142
.

3
)


Ming

Zhang,

and

John

Fulcher,



Face

recognition

using

artificial

neural

network

group
-
based

adaptive

tolerance

(GAT)

trees”,

IEEE

Transactions

on

Neural

Networkis
,

vol
.

7
,

no
.

3
,

pp
.

555
-
567
,

1996
.



2 Patents


1)


Ming Zhang, et al, “Translation invariant face recognition using network adaptive
tolerance tree”, Australian Patent PM 1828, Oct. 14, 1993.

2) Ming Zhang, Ruli Wang, and Yiming Gong, “Standard nonlinear signal wave generator
based on the neural network”, Chinese Patents, No. 90 1 02857.6, May 17, 1990.



Dr. Ming Zhang

Dr. Ming Zhang’ s Publications


(Face Recognition)

3 Full Refereed Conference Papers

1)


Shuxiang Xu, and Ming Zhang, A Novel Adaptive Activation Function, Accepted by IJCNN’2001
(International Joint Conference on Neural Networks’ 2001), Washington DC, USA, July 2001,
pp.2779


2782.

2)


Ming Zhang, Jing Chung Zhang, John Fulcher, "Neural network group models for data approximation",
International Journal of Neural Systems, Vol. 10, No. 2, April, 2000, pp. 123
-
142.

3
)


Ming

Zhang,

Shuxiang

Xu,

and

Bo

Lu,

“Neuron
-
adaptive

higher

order

neural

network

group

models”,

in

Proceedings

of

IJCNN’
99
,

Washington,

D
.
C
.
,

USA,

July

10
-
16
,

1999
.


4
)


Ming

Zhang,

Shuxiang

Xu,

Nigel

Bond,

and

Kate

Stevens,

“Neuron
-
adaptive

feedforward

neural

network

group

models”,

in

Proceedings

of

IASTED

International

Conference

on

Artificial

Intelligence

and

Soft

Computing
,

Honolulu,

Hawaii,

USA,

August

9
-
12
,

1999
,

pp
.
281
-
284
.

5
)


John

Fulcher,

Ming

Zhang,

“Translation
-
invariant

face

recognition

using

the

parellel

NAT
-
tree

neural

network

model”,

in

Proceedings

of

Parallel

ComputingWorkshop

1997
,

Canberra,

Australia,

25
-
26

September,

1997
,

pp
.

P
1
-
U
-
1



P
1
-
U
-
1
-
4
.

6
)


Ming

Zhang,

John

Fulcher,

“Face

recognition

system

using

NAT

tree”,

in

Proceedings

of

IASTED

International

Conference

on

Artificial

Intelligence

and

Soft

Computing
,

Banff,

Canada,

July

27

-

August

1
,

1997
,

pp
.

244
-
247
.

7
)


Ming

Zhang,

and

John

Fulcher,

“Face

perspective

understanding

using

artificial

neural

network

group
-
based

tree”,

in

Proceedings

of

International

Conference

on

Image

Processing
,

Lausanne,

Switzerland,

vol

III,

September

16
-
19
,

1996
,

pp
.
475
-
478
.

8
)

Ming

Zhang,

and

John

Fulcher,

“Translation

invariant

face

recognition

using

a

network

adaptive

tolerance

tree”,

in

Proceedings

of

World

Congress

On

Neural

Networks
,

San

Diego,

California,

USA,

September

15

-
18
,

1996
,

pp



Dr. Ming Zhang

Dr. Ming Zhang’ s Publications


Year 2001


(
1
)


Hui

Qi,

Ming

Zhang,

and

Roderick

Scofield,

Rainfall

Estimation

Using

M
-
PHONN

Model,

Accepted

by

IJCNN’
2001

(International

Joint

Conference

on

Neural

Networks’

2001
),

Washington

DC,

USA,

July

2001
,

pp
.

1620

-

1624
.

(
2
)


Ming

Zhang,

and

Roderick

Scofield,

Rainfall

Estimation

Using

A
-
PHONN

Model,

Accepted

by

IJCNN’
2001

(International

Joint

Conference

on

Neural

Networks’

2001
),

Washington

DC,

USA,

July

2001
,

pp
.

1583

-

1587
.

(
3
)


Ming

Zhang,

and

BO

Lu,

Financial

Data

Simulation

Using

M
-
PHONN

Model,

Accepted

by

IJCNN’
2001

(International

Joint

Conference

on

Neural

Networks’

2001
),

Washington

DC,

USA,

July

2001
,

pp
.

1828

-

1832
.

(
4
)


Ming

Zhang,

Financial

Data

Simulation

Using

A
-
PHONN

Model,

Accepted

by

IJCNN’
2001

(International

Joint

Conference

on

Neural

Networks’

2001
),

Washington

DC,

USA,

July

2001
,

pp
.
1823

-

1827
.

(
5
)


Shuxiang

Xu,

and

Ming

Zhang,

A

Novel

Adaptive

Activation

Function,

Accepted

by

IJCNN’
2001

(International

Joint

Conference

on

Neural

Networks’

2001
),

Washington

DC,

USA,

July

2001
,

pp
.
2779



2782


(
6
)

Ming

Zhang,

Rex

Gantenbein,

Sung

Y
.

Shin,

and

Chih
-
Cheng

Hung,

The

application

of



artificial

neural

networks

in

knowledge
-
based

information

systems,

International

Journal



of

Computer

and

Information

Science,

Vol

2
,

No
.
2
,

2001
,

pp
.
49

-

58
.

(
7
)


Ming

Zhang,

Shuxiang

Xu,

and

John

Fulcher,

Neuron
-
Adaptive

Higher

Order

Neural

Network

Models



for

Automated

Financial

Data

Modeling”,

Accepted

by

IEEEE

transactions

on

Neural

Networks,

July,



2001
.

Total

102

papers

published



Dr. Ming Zhang

Why This Project?

1.
Visual Studio.NET

2.
Image processing library

3.
Image capture source codes

4.
New generation computer models and techniques

5.
Plenty of research topics

6.
Good support of software and hardware

7.
Strong support from our Department

8.
Experienced supervisor

9.
Paper to be published in the International Conference

10.
Big market


Dr. Ming Zhang






PT
-
HONN Data Simulator




A
rtificial
N
eural network expert
S
ystem for
E
stimation of
R
ainfall from the satellite data


ANSER System (1991
-
2000)


-

1991
-
1992:US$66,000 suported by USA
National Research Council & NOAA

-

1995
-
1996:A$11,000 suppouted by Australia
Research Council& NOAA

-

1999
-
2000:US$62,000 suported by USA
National Research Council & NOAA





Why Develop ANSER ?


-

More than $3.5 billion in property is damaged
and, more than 225 people are killed by
heavy rain and flooding each year

-

No rainfall estimating system in GIS system,
No real time and working system of rainfall
estimation in the world

-

Can ANN be used in the weather forecasting
area? If yes, how should we use ANN
techniques in this area?





Why Use Neural Network Techniques ?


-

Two Directions of New generation computer


Quamtun Computer


Artificial Neural Network

-

Much quicker speed ?

-

Complicated pattern recognition?

-

Unknown rule knowledge base?

-

Self learning reasoning network?

-

Super position for multip choice?







ANSER Rainfall Estimation Result


9th May 2000

Time: 18Z



LAT LAN

Min 37.032 87.906

Max 38.765 88.480


ANSER

Min: 1.47 mm

Max: 6.37mm


NAVY

Min: 2.0mm

Max: 6.0mm

Conclusion
-

What Approved

Artificial Neural Network Techniques can :


-

Much quick speed: 5
-
10 time quick

-

Unknown rule knowledge base: Rainfall

-

Reasoning network: rainfall estimation


Conclusion
-

Next Step

-

Rebuild

interface

&

retraining

neural

networks

-

New

neural

netowrk

models
:



more complicated pattern recognition

-

Self expending knowledge base:


attract knowledge from real time cases

-

Self learning reasoning network: automatic system to

-

Study

in

advance

in

15

years
:

Artificial

Neural

Network

-

one

of

two

directions

of

new

generation

computer

Research




PHONN Simulator (1994
-

1996)

-

P
olynomial
H
igher
O
rder
N
eural
N
etwork financial data
simulator


-

A$ 105,000 Supported by Fujitsu, Japan


THONN Simulator (1996
-

1998)

-

T
rigonometric polynomial
H
igher
O
rder
N
eural
N
etwork
financial data simulator


-

A$ 10,000 Supported by Australia Research Council


PT
-
HONN Simulator (1999
-

2000)

-

P
olynomial and
T
rigonometric polynomial
H
igher
O
rder
N
eural
N
etwork financial data simulator


-

US$ 46,000 Supported by USA National Research Council








PT
-
HONN Data Simulator



Why Develop HONN ?


-
No system can automatically simulate
discontinue, unsmooth data very well

-
No system can automatically find the perfect
models for the discontinue, unsmooth data




Cloud Merge Using ANN Circle Operator











CONCLUSION
-

What Approved


The results of the comparative experiments show
that
THONG
system is able to simulate higher
frequency and higher order non
-
linear data, as well
as being able to simulate discontinuous data.


The
THONG
model can not only be used for
financial simulation, but also for financial
prediction.


Complicated pattern recognition: cloud merger





Conclusion
-

Next Step

-

Rebuild

interface

&

retraining

neural

networks


-

New

neural

network

models
:



more complicated pattern recognition


-
Financial data simulation experiments


-
Rainfall data simulation experiments