Английская грамматика
<<  Object-Oriented Software Engineering Practical Software Development using UML and Java ТЕОРЕТИЧЕСКАЯ ГРАММАТИКА  >>
Grouplet: A Structured Image Representation for Recognizing Human and
Grouplet: A Structured Image Representation for Recognizing Human and
Human-Object Interaction
Human-Object Interaction
Human-Object Interaction
Human-Object Interaction
Background: Human-Object Interaction
Background: Human-Object Interaction
Background: Human-Object Interaction
Background: Human-Object Interaction
Outline
Outline
Outline
Outline
Recognizing Human-Object Interaction is Challenging
Recognizing Human-Object Interaction is Challenging
Grouplet: our intuition
Grouplet: our intuition
Grouplet: our intuition
Grouplet: our intuition
Outline
Outline
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation (e
Grouplet representation
Grouplet representation
Grouplet representation
Grouplet representation
Outline
Outline
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
A “Space” of Grouplets
We only need discriminative Grouplets
We only need discriminative Grouplets
Obtaining discriminative grouplets for a class
Obtaining discriminative grouplets for a class
Using Grouplets for Classification
Using Grouplets for Classification
Outline
Outline
People-Playing-Musical-Instruments (PPMI) Dataset
People-Playing-Musical-Instruments (PPMI) Dataset
Recognition Tasks on People-Playing-Musical-Instruments (PPMI) Dataset
Recognition Tasks on People-Playing-Musical-Instruments (PPMI) Dataset
Classification: Playing Different Instruments
Classification: Playing Different Instruments
Classifying Playing vs
Classifying Playing vs
Classifying Playing vs
Classifying Playing vs
Detecting people playing musical instruments
Detecting people playing musical instruments
Detecting people playing musical instruments
Detecting people playing musical instruments
Detecting people playing musical instruments
Detecting people playing musical instruments
Examples of Mined Grouplets
Examples of Mined Grouplets
Conclusion
Conclusion
Thanks to
Thanks to

Презентация: «Видео про космос 2 класс». Автор: Bangpeng. Файл: «Видео про космос 2 класс.ppt». Размер zip-архива: 4547 КБ.

Видео про космос 2 класс

содержание презентации «Видео про космос 2 класс.ppt»
СлайдТекст
1 Grouplet: A Structured Image Representation for Recognizing Human and

Grouplet: A Structured Image Representation for Recognizing Human and

Object Interactions

Bangpeng Yao and Li Fei-Fei Computer Science Department, Stanford University

{bangpeng,feifeili}@cs.stanford.edu

1

2 Human-Object Interaction

Human-Object Interaction

Playing saxophone

Human

Not playing saxophone

Saxophone

2

3 Human-Object Interaction

Human-Object Interaction

Robots interact with objects

Automatic sports commentary

Medical care

“Kobe is dunking the ball.”

3

4 Background: Human-Object Interaction

Background: Human-Object Interaction

To be done

context

Schneiderman & Kanade, 2000 Viola & Jones, 2001 Huang et al, 2007 Papageorgiou & Poggio, 2000 Wu & Nevatia, 2005 Dalal & Triggs, 2005 Mikolajczyk et al, 2005 Leibe et al, 2005 Bourdev & Malik, 2009 Felzenszwalb & Huttenlocher, 2005 Ren et al, 2005 Ramanan, 2006 Ferrari et al, 2008 Yang & Mori, 2008 Andriluka et al, 2009 Eichner & Ferrari, 2009

Lowe, 1999 Belongie et al, 2002 Fergus et al, 2003 Fei-Fei et al, 2004 Berg & Malik, 2005 Felzenszwalb et al, 2005 Grauman & Darrell, 2005 Sivic et al, 2005 Lazebnik et al, 2006 Zhang et al, 2006 Savarese et al, 2007 Lampert et al, 2008 Desai et al, 2009 Gehler & Nowozin, 2009

Gupta et al, 2009

Yao & Fei-Fei, 2010a

Yao & Fei-Fei, 2010b

Murphy et al, 2003 Hoiem et al, 2006 Shotton et al, 2006

Rabinovich et al, 2007 Heitz & Koller, 2008 Divvala et al, 2009

vs.

4

5 Background: Human-Object Interaction

Background: Human-Object Interaction

To be done

context

Schneiderman & Kanade, 2000 Viola & Jones, 2001 Huang et al, 2007 Papageorgiou & Poggio, 2000 Wu & Nevatia, 2005 Dalal & Triggs, 2005 Mikolajczyk et al, 2005 Leibe et al, 2005 Bourdev & Malik, 2009 Felzenszwalb & Huttenlocher, 2005 Ren et al, 2005 Ramanan, 2006 Ferrari et al, 2008 Yang & Mori, 2008 Andriluka et al, 2009 Eichner & Ferrari, 2009

Lowe, 1999 Belongie et al, 2002 Fergus et al, 2003 Fei-Fei et al, 2004 Berg & Malik, 2005 Felzenszwalb et al, 2005 Grauman & Darrell, 2005 Sivic et al, 2005 Lazebnik et al, 2006 Zhang et al, 2006 Savarese et al, 2007 Lampert et al, 2008 Desai et al, 2009 Gehler & Nowozin, 2009

Gupta et al, 2009

Yao & Fei-Fei, 2010a

Yao & Fei-Fei, 2010b

Murphy et al, 2003 Hoiem et al, 2006 Shotton et al, 2006

Rabinovich et al, 2007 Heitz & Koller, 2008 Divvala et al, 2009

vs.

5

6 Outline

Outline

Intuition of Grouplet Representation Grouplet Feature Representation Using Grouplet for Recognition Dataset & Experiments Conclusion

6

7 Outline

Outline

Intuition of Grouplet Representation Grouplet Feature Representation Using Grouplet for Recognition Dataset & Experiments Conclusion

7

8 Recognizing Human-Object Interaction is Challenging

Recognizing Human-Object Interaction is Challenging

Reference image: playing saxophone

Different pose (or viewpoint)

Different lighting

Different background

Different instrument, similar pose

Same object (saxophone), different interactions

8

9 Grouplet: our intuition

Grouplet: our intuition

Bag-of-words

Spatial pyramid

Part-based

Grouplet Representation:

Thomas & Malik, 2001 Csurka et al, 2004 Fei-Fei & Perona, 2005 Sivic et al, 2005

Grauman & Darrell, 2005 Lazebnik et al, 2006

Weber et al, 2000 Fergus et al, 2003 Leibe et al, 2004 Felzenszwalb et al, 2005 Bourdev & Malik, 2009

9

10 Grouplet: our intuition

Grouplet: our intuition

Capture the subtle difference in human-object interactions.

Part-based configuration Co-occurrence Discriminative Dense

Grouplet Representation:

10

11 Outline

Outline

Intuition of Grouplet Representation Grouplet Feature Representation Using Grouplet for Recognition Dataset & Experiments Conclusion

11

12 Grouplet representation (e

Grouplet representation (e

g. 2-Grouplet)

Gaussian distribution

Visual codewords

Notations

I: Image. P: Reference point in the image. ?: Grouplet. ?i: Feature unit.

Ai: Visual codeword; xi: Image location; ?i: Variance of spatial distribution.

12

13 Grouplet representation (e

Grouplet representation (e

g. 2-Grouplet)

Gaussian distribution

Visual codewords

Notations

I: Image. P: Reference point in the image. ?: Grouplet. ?i: Feature unit. ?(?,I): Matching score of ? and I. ?(?i,I): Matching score of ?i and I.

Matching score between ? and I

Matching score between ?i and I

Ai: Visual codeword; xi: Image location; ?i: Variance of spatial distribution.

13

14 Grouplet representation (e

Grouplet representation (e

g. 2-Grouplet)

Gaussian distribution

Visual codewords

Notations

I: Image. P: Reference point in the image. ?: Grouplet. ?i: Feature unit. ?(?,I): Matching score of ? and I. ?(?i,I): Matching score of ?i and I. For an image patch: ?(x): Image neighborhood of x.

Matching score between ? and I

Matching score between ?i and I

Codeword assignment score

Gaussian density value

Ai: Visual codeword; xi: Image location; ?i: Variance of spatial distribution.

a?: Its visual appearance; x?: Its image location.

14

15 Grouplet representation (e

Grouplet representation (e

g. 2-Grouplet)

Gaussian distribution

Visual codewords

Notations

I: Image. P: Reference point in the image. ?: Grouplet. ?i: Feature unit. ?(?,I): Matching score of ? and I. ?(?i,I): Matching score of ?i and I. For an image patch: ?(x): Image neighborhood of x. ?: A small shift of the location.

Matching score between ? and I

Matching score between ?i and I

Codeword assignment score

Gaussian density value

Codeword assignment score

Gaussian density value

Ai: Visual codeword; xi: Image location; ?i: Variance of spatial distribution.

a?: Its visual appearance; x?: Its image location.

15

16 Grouplet representation

Grouplet representation

Playing saxophone

Other interactions

Part-based configuration Co-occurrence Discriminative

matching score: 0.6

matching score: 0.4

matching score: 0.0

matching score: 0.1

16

17 Grouplet representation

Grouplet representation

Part-based configuration Co-occurrence Discriminative Dense

All possible combinations of feature units

Densely sample image locations

Many possible spatial distributions

All possible Codewords

1-grouplet

2-grouplet

3-grouplet

17

18 Outline

Outline

Intuition of Grouplet Representation Grouplet Feature Representation Using Grouplet for Recognition Dataset & Experiments Conclusion

18

19 A “Space” of Grouplets

A “Space” of Grouplets

19

20 A “Space” of Grouplets

A “Space” of Grouplets

20

21 A “Space” of Grouplets

A “Space” of Grouplets

21

22 A “Space” of Grouplets

A “Space” of Grouplets

On background

Shared by different interactions

22

23 We only need discriminative Grouplets

We only need discriminative Grouplets

Number of feature units: N. N is large (192200)

Number of Grouplets: 2N very large space

On background

Shared by different interactions

Large ?(?,I)

Small ?(?,I)

Large ?(?,I)

Small ?(?,I)

23

23

24 Obtaining discriminative grouplets for a class

Obtaining discriminative grouplets for a class

Apriori Mining

Mine 1000~2000 grouplets, only need to evaluate (2~100)?N grouplets

Obtain grouplets with large ?(?,I) on the class.

Remove grouplets with large ?(?,I) from other classes.

Number of feature units: N. N is large (192200)

Number of Grouplets: 2N very large space

Selected 1-grouplets

Candidate 2-grouplets

[Agrawal & Srikant, 1994]

24

25 Using Grouplets for Classification

Using Grouplets for Classification

SVM

Discriminative grouplets

25

26 Outline

Outline

Intuition of Grouplet Representation Grouplet Feature Representation Using Grouplet for Recognition Dataset & Experiments Conclusion

26

27 People-Playing-Musical-Instruments (PPMI) Dataset

People-Playing-Musical-Instruments (PPMI) Dataset

http://vision.stanford.edu/resources_links.html

PPMI+

PPMI-

Normalized image (200 images each interaction)

Original image

# Image:

# Image:

(172)

(191)

(177)

(179)

(200)

(198)

(185)

(133)

(149)

(188)

(167)

(148)

(169)

(164)

27

28 Recognition Tasks on People-Playing-Musical-Instruments (PPMI) Dataset

Recognition Tasks on People-Playing-Musical-Instruments (PPMI) Dataset

Classification

Detection

Playing different instruments

Playing vs. Not playing

For each interaction, 100 training and 100 testing images.

vs.

vs.

28

29 Classification: Playing Different Instruments

Classification: Playing Different Instruments

7-class classification on PPMI+ images

SPM: [Lazebnik et al, 2006] DPM: [Felzenszwalb et al, 2008] Constellation: [Fergus et al, 2003] [Niebles & Fei-Fei, 2007]

29

30 Classifying Playing vs

Classifying Playing vs

Not playing

Seven 2-class classification problem; PPMI+ vs. PPMI- for each instrument.

Bassoon

Erhu

Flute

French horn

Saxophone

Violin

Average PPMI+ images

Average PPMI- images

30

31 Classifying Playing vs

Classifying Playing vs

Not playing

Seven 2-class classification problem; PPMI+ vs. PPMI- for each instrument.

Guitar

Average PPMI+ images

Average PPMI- images

31

32 Detecting people playing musical instruments

Detecting people playing musical instruments

Procedure:

Face detection with a low threshold; Crop and normalize image regions; 8-class classification

7 classes of playing instruments; Another class of not playing any instrument.

Playing saxophone

No playing

No playing

32

33 Detecting people playing musical instruments

Detecting people playing musical instruments

Area under the precision-recall curve:

Out method: 45.7%; Spatial pyramid: 37.3%.

33

34 Detecting people playing musical instruments

Detecting people playing musical instruments

Area under the precision-recall curve:

Out method: 45.7%; Spatial pyramid: 37.3%.

False detection

Missed detection

34

Playing French horn

35 Examples of Mined Grouplets

Examples of Mined Grouplets

Playing bassoon:

Playing saxophone:

Playing violin:

Playing guitar:

35

36 Conclusion

Conclusion

The Next Talk

Holistic image-based classification

Detailed understanding and reasoning

Vs.

Pose estimation & object detection

[B. Yao and L. Fei-Fei. “Grouplet: A structured image representation for recognizing human and object interactions.” CVPR 2010.]

[B. Yao and L. Fei-Fei. “Modeling mutual context of object and human pose in human-object interaction activities.” CVPR 2010.]

36

37 Thanks to

Thanks to

Juan Carlos Niebles, Jia Deng, Jia Li, Hao Su, Silvio Savarese, and anonymous reviewers. And You

37

«Видео про космос 2 класс»
http://900igr.net/prezentacija/anglijskij-jazyk/video-pro-kosmos-2-klass-255685.html
cсылка на страницу
Урок

Английский язык

29 тем
Слайды