cs 1699: intro to computer vision detection ii: deformable part models prof. adriana kovashka...
TRANSCRIPT
![Page 1: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/1.jpg)
CS 1699: Intro to Computer Vision
Detection II: Deformable Part Models
Prof. Adriana KovashkaUniversity of Pittsburgh
November 12, 2015
![Page 2: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/2.jpg)
Today: Object category detection
• Window-based approaches:– Review: Viola-Jones detector– Dalal-Triggs pedestrian detector
• Part-based approaches:– Implicit shape model– Deformable parts model
![Page 3: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/3.jpg)
Object Category Detection• Focus on object search: “Where is it?”• Build templates that quickly differentiate object
patch from background patch
Dog or Non-Dog?
…
Derek Hoiem
![Page 4: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/4.jpg)
Viola-Jones detector: features
Feature output is difference between adjacent regions
Efficiently computable with integral image: any sum can be computed in constant time
“Rectangular” filters
Value at (x,y) is sum of pixels above and to the left of (x,y)
Integral image
Adapted from Kristen Grauman and Lana Lazebnik
Value = ∑ (pixels in white area) – ∑ (pixels in black area)
![Page 5: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/5.jpg)
Considering all possible filter parameters: position, scale, and type:
180,000+ possible features associated with each 24 x 24 window
Which subset of these features should we use to determine if a window has a face?
Use AdaBoost both to select the informative features and to form the classifier
Viola-Jones detector: features
Kristen Grauman
![Page 6: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/6.jpg)
Viola-Jones detector: AdaBoost• Want to select the single rectangle feature and
threshold that best separates positive (faces) and negative (non-faces) training examples, in terms of weighted error.
Outputs of a possible rectangle feature on faces and non-faces.
…
Resulting weak classifier:
For next round, reweight the examples according to errors, choose another filter/threshold combo.
Kristen Grauman
![Page 7: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/7.jpg)
Cascading classifiers for detection
• Form a cascade with low false negative rates early on
• Apply less accurate but faster classifiers first to immediately discard windows that clearly appear to be negative
Kristen Grauman
DEMO
![Page 8: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/8.jpg)
Dalal-Triggs pedestrian detector
1. Extract fixed-sized (64x128 pixel) window at each position and scale
2. Compute HOG (histogram of gradient) features within each window
3. Score the window with a linear SVM classifier4. Perform non-maxima suppression to remove
overlapping detections with lower scoresNavneet Dalal and Bill Triggs, Histograms of Oriented Gradients for Human Detection, CVPR05
![Page 9: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/9.jpg)
Resolving detection scores
Non-max suppression
Score = 0.1
Score = 0.8 Score = 0.8
Adapted from Derek Hoiem
![Page 10: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/10.jpg)
Slide by Pete Barnum, Kristen Grauman Navneet Dalal and Bill Triggs, Histograms of Oriented Gradients for Human Detection, CVPR05
•Map each grid cell in the input window to a histogram counting the gradients per orientation.
•Train a linear SVM using training set of pedestrian vs. non-pedestrian windows.
Code available: http://pascal.inrialpes.fr/soft/olt/
![Page 11: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/11.jpg)
• Histogram of gradient orientations
– Votes weighted by magnitude– Bilinear interpolation between cells
Orientation: 9 bins (for unsigned angles)
Histograms in 8x8 pixel cells
Slide by Pete Barnum Navneet Dalal and Bill Triggs, Histograms of Oriented Gradients for Human Detection, CVPR05
![Page 12: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/12.jpg)
Histograms of oriented gradients (HOG)
N. Dalal and B. Triggs, Histograms of Oriented Gradients for Human Detection, CVPR 2005
10x10 cells
20x20 cells
Image credit: N. Snavely
![Page 13: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/13.jpg)
Histograms of oriented gradients (HOG)
N. Dalal and B. Triggs, Histograms of Oriented Gradients for Human Detection, CVPR 2005
Image credit: N. Snavely
![Page 14: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/14.jpg)
Histograms of oriented gradients (HOG)
N. Dalal and B. Triggs, Histograms of Oriented Gradients for Human Detection, CVPR 2005
![Page 15: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/15.jpg)
Histograms of oriented gradients (HOG)
N. Dalal and B. Triggs, Histograms of Oriented Gradients for Human Detection, CVPR 2005
![Page 16: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/16.jpg)
Slide by Pete Barnum Navneet Dalal and Bill Triggs, Histograms of Oriented Gradients for Human Detection, CVPR05
pos w neg w
![Page 17: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/17.jpg)
pedestrian
Slide by Pete Barnum Navneet Dalal and Bill Triggs, Histograms of Oriented Gradients for Human Detection, CVPR05
![Page 18: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/18.jpg)
Detection examples
![Page 19: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/19.jpg)
Are we done?• Single rigid template usually not enough to
represent a category– Many objects (e.g. humans) are articulated, or have parts
that can vary in configuration
– Many object categories look very different from different viewpoints, or from instance to instance
Slide by N. Snavely
![Page 20: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/20.jpg)
Deformable objects
Images from Caltech-256
Slide Credit: Duan Tran
![Page 21: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/21.jpg)
Deformable objects
Images from D. Ramanan’s datasetSlide Credit: Duan Tran
![Page 22: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/22.jpg)
Parts-based Models
Define object by collection of parts modeled by1. Appearance2. Spatial configuration
Slide credit: Rob Fergus
![Page 23: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/23.jpg)
How to model spatial relations?• One extreme: fixed template
Derek Hoiem
![Page 24: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/24.jpg)
Part-based template
• Object model = sum of scores of features at fixed positions
+3 +2 -2 -1 -2.5 = -0.5
+4 +1 +0.5 +3 +0.5 = 10.5
> 7.5?
> 7.5?
Non-object
Object
Derek Hoiem
![Page 25: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/25.jpg)
How to model spatial relations?• Another extreme: bag of words
=
Derek Hoiem
![Page 26: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/26.jpg)
How to model spatial relations?• Star-shaped model
Root
Part
Part
Part
Part
Part
Derek Hoiem
![Page 27: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/27.jpg)
How to model spatial relations?• Star-shaped model
=X X
X
≈Root
Part
Part
Part
Part
Part
Derek Hoiem
![Page 28: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/28.jpg)
Parts-based Models• Articulated parts model
– Object is configuration of parts– Each part is detectable and can move around
Adapted from Derek Hoiem, images from Felzenszwalb
![Page 29: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/29.jpg)
Implicit shape models• Visual vocabulary is used to index votes for
object position [a visual word = “part”]
B. Leibe, A. Leonardis, and B. Schiele, Combined Object Categorization and Segmentation with an Implicit Shape Model, ECCV Workshop on Statistical Learning in Computer Vision 2004
visual codeword withdisplacement vectors
training image annotated with object localization info
Lana Lazebnik
![Page 30: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/30.jpg)
Implicit shape models• Visual vocabulary is used to index votes for
object position [a visual word = “part”]
B. Leibe, A. Leonardis, and B. Schiele, Combined Object Categorization and Segmentation with an Implicit Shape Model, ECCV Workshop on Statistical Learning in Computer Vision 2004
test image
Lana Lazebnik
![Page 31: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/31.jpg)
Implicit shape models: Training
1. Build vocabulary of patches around extracted interest points using clustering
Lana Lazebnik
![Page 32: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/32.jpg)
Implicit shape models: Training
1. Build vocabulary of patches around extracted interest points using clustering
2. Map the patch around each interest point to closest word
Lana Lazebnik
![Page 33: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/33.jpg)
Implicit shape models: Training
1. Build vocabulary of patches around extracted interest points using clustering
2. Map the patch around each interest point to closest word
3. For each word, store all positions it was found, relative to object center
Lana Lazebnik
![Page 34: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/34.jpg)
Implicit shape models: Testing1. Given new test image, extract patches, match to
vocabulary words
2. Cast votes for possible positions of object center
3. Search for maxima in voting space
Lana Lazebnik
![Page 35: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/35.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
Original image
Example: Results on Cows
![Page 36: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/36.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
Original image
Interest points
Example: Results on Cows
![Page 37: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/37.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
Original image
Interest points
Matched patches
Example: Results on Cows
![Page 38: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/38.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
1st hypothesis
Example: Results on Cows
![Page 39: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/39.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
2nd hypothesis
Example: Results on Cows
![Page 40: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/40.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
Example: Results on Cows
3rd hypothesis
![Page 41: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/41.jpg)
Perc
ep
tual an
d S
en
sory
Au
gm
en
ted
Com
pu
tin
gV
isu
al O
bje
ct
Recog
nit
ion
Tu
tori
al
K. Grauman, B. Leibe
Detection Results
• Qualitative Performance Recognizes different kinds of objects Robust to clutter, occlusion, noise, low contrast
![Page 42: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/42.jpg)
Discriminative part-based models
P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan, Object Detection with Discriminatively Trained Part Based Models, PAMI
32(9), 2010
Root filter
Part filters
Deformation weights
Lana Lazebnik
![Page 43: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/43.jpg)
Discriminative part-based models
P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan, Object Detection with Discriminatively Trained Part Based Models, PAMI
32(9), 2010
Multiple components
Lana Lazebnik
![Page 44: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/44.jpg)
Discriminative part-based models
P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan, Object Detection with Discriminatively Trained Part Based Models, PAMI
32(9), 2010 Lana Lazebnik
![Page 45: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/45.jpg)
Scoring an object hypothesis• The score of a hypothesis is
the sum of appearance scores minus the sum of deformation costs
),,,()(),...,( 22
0 10 ii
n
i
n
iiiiiin dydxdydxscore
DpHFpp
Appearance weights
Subwindow features
Deformation weights
Displacements
Adapted from Lana Lazebnik
![Page 46: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/46.jpg)
Detection
Lana Lazebnik
![Page 47: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/47.jpg)
Training• Training data consists of images with labeled
bounding boxes• Need to learn the filters and deformation parameters
Lana Lazebnik
![Page 48: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/48.jpg)
Training• Our classifier has the form
• w are model parameters, z are latent hypotheses
• Latent SVM training:• Initialize w and iterate:
• Fix w and find the best z for each training example • Fix z and solve for w (standard SVM training)
),(max)( zxHwx z f
Lana Lazebnik
![Page 49: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/49.jpg)
Car model
Component 1
Component 2
Lana Lazebnik
![Page 50: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/50.jpg)
Car detections
Lana Lazebnik
![Page 51: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/51.jpg)
Person model
Lana Lazebnik
![Page 52: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/52.jpg)
Person detections
Lana Lazebnik
![Page 53: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/53.jpg)
Bottle model
Lana Lazebnik
![Page 54: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/54.jpg)
Cat model
Lana Lazebnik
![Page 55: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/55.jpg)
Cat detections
Lana Lazebnik
![Page 56: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/56.jpg)
Detection state of the art
Object detection system overview. Our system (1) takes an input image, (2) extracts around 2000 bottom-up region proposals, (3) computes features for each proposal using a large convolutional neural network (CNN), and then (4) classifies each region using class-specific linear SVMs. R-CNN achieves a mean average precision (mAP) of 53.7% on PASCAL VOC 2010. For comparison, Uijlings et al. (2013) report 35.1% mAP using the same region proposals, but with a spatial pyramid and bag-of-visual-words approach. The popular deformable part models perform at 33.4%.
R. Girshick, J. Donahue, T. Darrell, and J. Malik, Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation, CVPR 2014.
Lana Lazebnik
![Page 57: CS 1699: Intro to Computer Vision Detection II: Deformable Part Models Prof. Adriana Kovashka University of Pittsburgh November 12, 2015](https://reader035.vdocuments.mx/reader035/viewer/2022062805/5697bfb51a28abf838c9d839/html5/thumbnails/57.jpg)
Summary• Window-based approaches
– Assume object appears in roughly the same configuration in different images
– Look for alignment with a global template• Part-based methods
– Allow parts to move somewhat from their usual locations
– Look for good fits in appearance, for both the global template and the individual part templates
• Next time: Analyzing and debugging detection methods