class-specific hough forests for object detection
DESCRIPTION
Class-Specific Hough Forests for Object Detection. Zhen Yuan Hsu Advisor : S.J.Wang. Gall, J., Lempitsky , V.: Class- specic hough forests for object detection. In: IEEE CVPR(2009 ). Outline. Related work Why we use Random forest What’s Hough forest - PowerPoint PPT PresentationTRANSCRIPT
Class-Specific Hough Forestsfor Object Detection
Zhen Yuan HsuAdvisor : S.J.Wang
Gall, J., Lempitsky, V.: Class-specic hough forests for object detection. In: IEEE CVPR(2009)
Outline
• Related work• Why we use Random forest• What’s Hough forest• How Hough forest work for object detection
Implicit shape models: Training• Extract 25x25 patches around Harris corners.• Generate a codebook of local appearance patches using
clustering.• For each cluster, extract its center and store it in the
codebook.• For each codebook entry, store all positions it was found
relative to object center.
Implicit shape models: Testing1. Given test image, extract patches, match to codebook
entry 2. Cast votes for possible positions of object center3. Search for maxima in voting space4. Extract weighted segmentation mask based on stored
masks for the codebook occurrencesMatch 、 offset
Why we use Random forest
Time 、 Training data
Random forest
Decision tree
x1>w1
x2>w2
Yes
Yes
No
No
x1
x2
W1
W2
A Forest
……tree t1 tree tT
category ccategory c
split nodesleaf nodes
v v
What’s RandomnessRandomness – Data and Split fuction
for each node :Split fuction is randomly selected.
Binary Tests
• selected during training from a random subset of all split functions.
split node
. P .q
a threshold
: 16*16 image feature
choice
Randomness - Split fuction• Try several lines,
chosen at random
• Keep line that best separates data– information gain
• Recurse
Random forest for object detection
Object localization x : regression
Classfying patch belong to object c :classification
datax
y
What’s Hough forest
Random forest Hough vote
Hough forest
Hough Forests: Training• Supervised learning
• Label:negative or background samples (blue)positive samples (red)offset vectors (green)
Feature of local patch
Hough Forests: Training
……
split nodesleaf nodes
CL : positive sample patch proportion
Leavestwo important information for voting:1.CL : positive sample patch proportion2. DL={di} , iϵA
Stop criteriaLeaf condition : 1. number of image patches < ϵ 2.a threshold based on minimum of uncertainty(Class-label , Offset vector)
Quality of Binary Tests• Goal :Minimize the Class-label uncertainty and Offset uncertainty:
• Type of uncertainty is randomly selected for each node
• Class-label uncertainty:
• Offset uncertainty:
A=the set of all image patch={ }Ci=class label
Detection
Original imageInterest pointsMatched patches
Position y .
Detection
……
Possible Center of objet : y+di
1.CL : positive sample patch proportion2.DL={di} iϵA
Position y .
Hough vote
Probabilistic votesSource: B. Leibe
Position y .
d2
d1d3
Hough voteFor location x and given image patch I(y) and tree T
x : center of bounding box x≈y+di
• Confidence vote : 1.CL =weight 2. di :offest vector
Over all trees:
Accumulation over all image patches:
Detection
Multi-Scale and Multi-Ratio• Multi Scale: 3D Votes (x, y, scale)
• Multi-Ratio: 4D Votes (x, y, scale, ratio)
UIUC Cars - Multi Scale
• Wrong (EER)
• Correct
Comparison
Pedestrians (INRIA)
Pedestrians (INRIA)
Pedestrians (TUD)
reference• http://mi.eng.cam.ac.uk/~tkk22/iccv09_tutorial• 利用霍夫森林建構行人偵測技術 - 清華電機系 陳仕儒碩士論文 2012• An Introduction to Random Forests for Multi-class Object Detection, J.Gall
• Thank you for your listening!