Groups of Adjacent Contour Segments for Object Detection

1
Groups of Adjacent Contour Segments for Object
Detection

• Vittorio Ferrari
• Loic Fevrier
• Frederic Jurie
• Cordelia Schmid

2
Problem object class detection localization
Training
Focus classes with characteristic shape
Testing
3
Features pairs of adjacent segments (PAS)

• Contour segment network
• Ferrari et al. ECCV 2006
• edgels extracted with Berkeley boundary detector
• 2) edgel-chains partitioned into straight
  contour segments
• 3) segments connected at edgel-chains endpoints
  and junctions

4
Features pairs of adjacent segments (PAS)
segments connected in the network
PAS groups of two connected segments

• encodes geometric properties of the PAS
• scale and translation invariant
• compact, 5D

5
Features pairs of adjacent segments (PAS)
intermediate complexity good
repeatability-informativeness trade-off
scale-translation invariant
connected natural grouping criterion (need not
choose a grouping neighborhood or scale)
6
PAS codebook
Based on descriptors, cluster PAS into types
a few of the most frequent types based on 10
outdoor images (5 horses and 5 background).
types based on 15 indoor images (bottles)

• Frequently occurring PAS have intuitive, natural
  shapes
• As we add images, number of PAS types converges
  to just 100
• Very similar codebooks come out, regardless of
  source images
• general, simple features. We use a single,
  universal codebook (1st row) for all classes

7
Window descriptor
1. Subdivide window into tiles. 2. Compute a
separate bag of PAS per tile 3. Concatenate these
semi-local bags Lazebnik et al. CVPR 2006
Dalal and Triggs CVPR 2005 distinctive
records which PAS appear where weight
PAS by average edge strength flexible
soft-assign PAS to types rather coarse
tiling fast to compute using Integral
Histograms
8
Training
1. Learn mean positive window dimensions 2.
Determine number of tiles T 3. Collect positive
example descriptors
9
Training
5. Train a linear SVM
Here a few of the top weighted descriptor vector
dimensions ( 'PAS tile')
lie on object boundary ( local shape structure
common to many training examples)
10
Testing
1. Slide window of aspect ratio
, at multiple scales
2. SVM classify each window non-maxima
suppression detections
11
Results INRIA horses
Dataset Jurie and Schmid, CVPR 2004
170 positive 170 negative images (training
50 pos 50 neg) wide range of
scales clutter
tiling brings a substantial improvement
optimum at T30 -gt keep this setting on all other
experiments
works well 86 det-rate at 0.3 FPPI (with 50
pos 50 neg training images)
12
Results INRIA horses
Dataset Jurie and Schmid, CVPR 2004
170 positive 170 negative images (training
50 pos 50 neg) wide range of
scales clutter
PAS better than any IP all interest point
(IP) comparisons with T10, and 120 feature
types, ( optimum over INRIA horses, and ETHZ
Shape Classes all IP codebooks are
class-specific)
13
Results Weizmann-Shotton horses
Dataset Shotton et al., ICCV 2005
327 positive 327 negative images (training
50 pos 50 neg) no scale changes
modest clutter
Shottons EER
- exact comparison to Shotton et al. use their
images and search at a single scale - PAS same
performance (92 precision-recall EER), but
no need for segmented training images (only
bounding-boxes) can detect objects at
multiple scales (see other experiments)
14
Results ETHZ Shape Classes
Dataset Ferrari et al., ECCV 2006
255 images, over 5 classes training
half of positive images for a class
same number from the other
classes (1/4 from each) testing
all other images large scale
changes extensive clutter
15
Results ETHZ Shape Classes
Dataset Ferrari et al., ECCV 2006
255 images, over 5 classes training
half of positive images for a class
same number from the other
classes (1/4 from each) testing
all other images large scale
changes extensive clutter
16
Results ETHZ Shape Classes
mean det-rate at 0.4 FPPI 79
class specific IP codebooks
PAS gtgt I.P for apple logos, bottles, mugs
PAS IP for giraffes (texture!)
PAS lt IP for swan
overall best IP Harris-Laplace
17
Results Caltech 101
Results Caltech 101
Dataset Fei-Fei et al., GMBV 2004
42 anchor, 62 chair, 67 cup images train half
same number of caltech101 background testing
other half pos same number of background scale
changes only little clutter
18
Results Caltech 101
Dataset Fei-Fei et al., GMBV 2004
On caltech101s anchor, chair, cup PAS better
than Harris-Laplace mean PAS det-rate at 0.4
FPPI 85
19
Comparison to Dalal and Triggs CVPR 2005
20
Comparison to Dalal and Triggs CVPR 2005
overall mean det-rate at 0.4 FPPI PAS 82 gtgt
HoG 58
PAS gtgt HoG for 6 datasets PAS HoG for 2
datasets PAS lt HoG for 2 datasets
21
Generalizing PAS to kAS
kAS any path of length k through the contour
segment network
segments connected in the network
4AS
3AS

• scaletranslation invariant descriptor with
  dimensionality 4k-2
• k feature complexity higher k -gt more
  informative, but less repeatable kAS
• overall mean det-rates ()
• 1AS PAS 3AS
  4AS
• 0.3 FPPI 69 77 64
  57
• 0.4 FPPI 76 82 70
  64

PAS do best !
22
Conclusions
Connected local shape features for object class
detection
Experiments on 10 diverse classes from 4 datasets
show
better suited than interest points for these
shape-based classes
PAS have the best intermediate complexity among
kAS
object detector deals with clutter, scale
changes, intra-class variability
object detector compares favorably to HoG-based
one
- fixed aspect-ratio window sometimes inaccurate
bounding-boxes
- single viewpoint
23
Current work detecting object outlines
Training learn the common boundaries from
examples
24
Current work detecting object outlines
Detection on a new image
1. detect edges
2. match PAS based on descriptors
25
A few preliminary results
26
Results Caltech 101
Dataset Fei-Fei et al., GMBV 2004
On caltech101s anchor, chair, cup PAS better
than any IP mean PAS det-rate at 0.4 FPPI 85