Download - Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context
![Page 1: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/1.jpg)
Context for single object classes
![Page 2: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/2.jpg)
Who needs context anyway?We can recognize objects even out of context
Banksy
![Page 3: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/3.jpg)
Why is context important?• Changes the interpretation of an object (or its function)
• Context defines what an unexpected event is
![Page 4: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/4.jpg)
Look-Alikes by Joan Steiner
Even in high resolution, we can not shut down contextual processing and it is hard to recognize the true identities of the elements that compose this scene.
![Page 5: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/5.jpg)
![Page 6: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/6.jpg)
![Page 7: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/7.jpg)
![Page 8: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/8.jpg)
![Page 9: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/9.jpg)
The importance of context• Cognitive psychology
– Palmer 1975 – Biederman 1981– …
• Computer vision– Noton and Stark (1971)– Hanson and Riseman (1978)– Barrow & Tenenbaum (1978) – Ohta, kanade, Skai (1978)– Haralick (1983)– Strat and Fischler (1991)– Bobick and Pinhanez (1995)– Campbell et al (1997)
![Page 10: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/10.jpg)
What is the context for a single object category?
![Page 11: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/11.jpg)
The influence of an object extends beyond its physical boundaries
![Page 12: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/12.jpg)
Global and local representationsbuilding
car
sidewalk
Urban street scene
![Page 13: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/13.jpg)
Global and local representations
Image index: Summary statistics, configuration of textures
Urban street scene
features
histogram
building
car
sidewalk
Urban street scene
![Page 14: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/14.jpg)
Global scene representations
Spatial structure is important in order to provide context for object localization
Sivic et. al., ICCV 2005Fei-Fei and Perona, CVPR 2005
Bag of words Spatially organized textures
Non localized textons
S. Lazebnik, et al, CVPR 2006Walker, Malik. Vision Research 2004 …
M. Gorkani, R. Picard, ICPR 1994A. Oliva, A. Torralba, IJCV 2001
…
![Page 15: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/15.jpg)
An integrated model of Scenes, Objects, and Parts
Ncar
S
g
Scene
Scenegist
features
0
0
1
1
5
5
N
P(Ncar | S = street)
P(Ncar | S = park)
0 5 10 150
0.05
0.1
0.15
0.2
0 5 10 150
0.2
0.4
0.6
0.8
N
![Page 16: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/16.jpg)
Context driven object detection
ZcarNcar
S
g
Scene
Scenegist
features
01 5
P(Ncar | S = street)
0 5 10 150
0.05
0.1
0.15
0.2
N
![Page 17: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/17.jpg)
An integrated model of Scenes, Objects, and Parts
p(d | F=1) = N(d | m1, s1)p(d | F=0) = N(d | m0, s0)
We train a multiview car detector.
xcari
dcari
carFi
N=4
![Page 18: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/18.jpg)
An integrated model of Scenes, Objects, and Parts
ZcarNcar
S
g
Scene
Scenegist
features
xcari
dcari
carFi
M=4
P(F,S | x,d,g) a p(F | S)p(S | g) p(xi | g) PN(xi; mb, sb2) PN(di; mtp, stp
2) PN(di; mtn, stn2)
i:Fi=0 i:Fi=1 i:Fi=0
![Page 19: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/19.jpg)
![Page 20: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/20.jpg)
![Page 21: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/21.jpg)
A car out of context …
![Page 22: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/22.jpg)
We are wired for 3D~6cm
![Page 23: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/23.jpg)
We can not shut down 3D perception
(c) 2006 Walt Anthony
![Page 24: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/24.jpg)
Scenes rule over objects
3D percept is driven by the scene, which imposes its ruling to the objects
![Page 25: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/25.jpg)
3D from pixel valuesD. Hoiem, A.A. Efros, and M. Hebert, "Automatic Photo Pop-up”. SIGGRAPH 2005.
A. Saxena, M. Sun, A. Y. Ng. "Learning 3-D Scene Structure from a Single Still Image"In ICCV workshop on 3D Representation for Recognition (3dRR-07), 2007.
![Page 26: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/26.jpg)
Surface EstimationImage Support Vertical Sky
V-Left V-Center V-Right V-Porous V-Solid
[Hoiem, Efros, Hebert ICCV 2005]
ObjectSurface?
Support?Slide by Derek Hoiem
![Page 27: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/27.jpg)
Object Support
Slide by Derek Hoiem
![Page 28: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/28.jpg)
Slide by James Coughlan
![Page 29: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/29.jpg)
Slide by James Coughlan
![Page 30: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/30.jpg)
3d Scene Context
Image World
Hoiem, Efros, Hebert ICCV 2005
![Page 31: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/31.jpg)
3D scene context
meters
met
ers
Ped
Ped
Car
Hoiem, Efros, Hebert ICCV 2005
![Page 32: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/32.jpg)
Qualitative Results
Initial: 2 TP / 3 FP Final: 7 TP / 4 FP
Local Detector from [Murphy-Torralba-Freeman 2003]
Car: TP / FP Ped: TP / FP
Slide by Derek Hoiem
![Page 33: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/33.jpg)
3D City Modeling using Cognitive Loops
N. Cornelis, B. Leibe, K. Cornelis, L. Van Gool. CVPR'06
![Page 34: Iccv2009 recognition and learning object categories p1 c02 - detecting single objects in context](https://reader033.vdocument.in/reader033/viewer/2022061213/54968cf1b4795923248b4573/html5/thumbnails/34.jpg)
Single view metrologyCriminisi, et al. 1999
Need to recover:• Ground plane• Reference height• Horizon line• Where objects contact the ground