SlideShare une entreprise Scribd logo
1  sur  36
Max-Margin Additive Classifiers for Detection SubhransuMaji & Alexander Berg University of California at Berkeley  Columbia University ICCV 2009, Kyoto, Japan
Accuracy vs. Evaluation Timefor SVM Classifiers Non-linear Kernel Evaluation time Linear Kernel Accuracy
Accuracy vs. Evaluation Timefor SVM Classifiers Non-linear Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy
Non-linear Kernel  Additive Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy Accuracy vs. Evaluation Timefor SVM Classifiers
Additive Kernel Non-linear Kernel  Additive Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy Accuracy vs. Evaluation Timefor SVM Classifiers
Accuracy vs. Evaluation Timefor SVM Classifiers Additive Kernel Non-linear Kernel Evaluation time Our CVPR 08 Linear Kernel  Additive Kernel Accuracy Made it possible to use SVMs with additive kernels for detection.
Additive Classifiers Much work already uses them!	 SVMs with additive kernels are additive classifiers Histogram based kernels Histogram intersection, chi-squared kernel Pyramid Match Kernel (Grauman & Darell, ICCV’05) Spatial Pyramid Match Kernel (Lazebniket.al., CVPR’06) ….
Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time Linear Kernel Accuracy
Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time <=1990s Linear Accuracy
Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time Today Linear Accuracy Eg. Cutting Plane, Stoc. Gradient Descend, Dual Coordinate Descend
Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time Our CVPR 08 Linear Accuracy
Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time Our CVPR 08 ✗ Linear Accuracy
Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time This Paper Linear Accuracy
Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time This Paper Linear Additive Accuracy Makes it possible to train additive classifiers very fast.
Summary Additive classifiers are widely used and can provide better accuracy than linear Our CVPR 08: SVMs with additive kernels are additive classifiers and can be evaluated in O(#dim) -- same as linear. This work:  additive classifiers can be trained directly as efficiently (up to a small constant) as the best approaches for training linear classifiers. An example
Support Vector Machines Embedded Space Input Space Kernel Function ,[object Object]
 Can learn non-linear boundaries in input space Classification Function Kernel Trick
Embeddings… These embeddings can be high dimensional (even infinite) Our approach is based on embeddings thatapproximate kernels. We’d like this to be as accurate as possible We are going to use fast linear classifier training algorithms on the             so sparseness is important.
Key Idea: Embedding an Additive Kernel Additive Kernels are easy to embed, just embed each dimension independently Linear Embedding for min Kernel for integers For non integers can approximate by quantizing
Issues: Embedding Error Quantization leads to large errors Better encoding x y
Issues: Sparsity Represent with sparse values
Linear SVM objective (solve with LIBLINEAR): Encoded SVM objective (not practical):  Linear vs. Encoded SVMs
Linear vs. Encoded SVMs  Linear SVM objective (solve with LIBLINEAR): Encoded SVM modified (custom solver):  Encourages smooth functions Closely approximates min kernel SVM Custom solver : PWLSGD (see paper)
Linear SVM objective (solve with LIBLINEAR): Encoded SVM objective (solve with LIBLINEAR) :  Linear vs. Encoded SVMs
Additive Classifier Choices Regularization Encoding
Additive Classifier Choices Accuracy Increases  Regularization Encoding Evaluation times are similar
Additive Classifier Choices Accuracy Increases  Regularization Encoding Accuracy Increases  Evaluation times are similar
Additive Classifier Choices Accuracy Increases  Regularization Encoding Accuracy Increases  Standard solver Eg. LIBSVM Few lines of code + standard solver Eg. LIBLINEAR
Additive Classifier Choices Accuracy Increases  Regularization Encoding Accuracy Increases  Custom solver
Additive Classifier Choices Accuracy Increases  Regularization Encoding Accuracy Increases  Classifier Notations
Experiments “Small” Scale: Caltech 101 (Fei-Fei, et.al.) “Medium” Scale: DC Pedestrians (Munder & Gavrila) “Large” Scale : INRIA Pedestrians (Dalal & Triggs)
Experiment : DC Pedestrians (3.18s, 89.25%) (1.86s, 88.80%) (363s, 89.05%) (2.98s, 85.71%) 100x faster training time ~ linear SVM accuracy ~ kernel SVM  (1.89s, 72.98%) 20,000 features, 656 dimensional 100 bins for encoding 6-fold cross validation
Experiment : Caltech 101 (291s, 55.35%) (2687s, 56.49%) (102s, 54.8%) (90s, 51.64%) 10x faster Small loss in accuracy (41s, 46.15%) 30 training examples per category 100 bins for encoding Pyramid HOG + Spatial Pyramid Match Kernel
Experiment : INRIA  Pedestrians (140 mins, 0.95) (76s, 0.94) (27s, 0.88) 300x faster training time ~ linear SVM accuracy ~ kernel SVMtrains the detector in < 2 mins  (122s, 0.85) (20s, 0.82) SPHOG: 39,000 features, 2268 dimensional  100 bins for encoding Cross Validation Plots
Experiment : INRIA  Pedestrians 300x faster training time ~ linear SVM accuracy ~ kernel SVMtrains the detector in < 2 mins  SPHOG: 39,000 features, 2268 dimensional  100 bins for encoding Cross Validation Plots
Take Home Messages Additive models are practical for large scale data Can be trained discriminatively:	 Poor man’s version : encode + Linear SVM Solver Middle man’s version : encode + Custom Solver Rich man’s version : Min Kernel SVM Embedding only Approximates kernels, leads to small loss in accuracy but up to 100x speedup in training time Everyone should use: see code on our websites Fast IKSVM from CVPR’08, Encoded SVMs, etc

Contenu connexe

Similaire à ICCV2009: Max-Margin Ađitive Classifiers for Detection

Data Science - Part IX - Support Vector Machine
Data Science - Part IX -  Support Vector MachineData Science - Part IX -  Support Vector Machine
Data Science - Part IX - Support Vector MachineDerek Kane
 
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...DataScienceConferenc1
 
Huong dan cu the svm
Huong dan cu the svmHuong dan cu the svm
Huong dan cu the svmtaikhoan262
 
Machine learning, biomarker accuracy and best practices
Machine learning, biomarker accuracy and best practicesMachine learning, biomarker accuracy and best practices
Machine learning, biomarker accuracy and best practicesPradeep Redddy Raamana
 
Optimizing Terascale Machine Learning Pipelines with Keystone ML
Optimizing Terascale Machine Learning Pipelines with Keystone MLOptimizing Terascale Machine Learning Pipelines with Keystone ML
Optimizing Terascale Machine Learning Pipelines with Keystone MLSpark Summit
 
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...Exploration of Supervised Machine Learning Techniques for Runtime Selection o...
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...Akihiro Hayashi
 
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...Lviv Data Science Summer School
 
4. Classification.pdf
4. Classification.pdf4. Classification.pdf
4. Classification.pdfJyoti Yadav
 
Types of Machine Learnig Algorithms(CART, ID3)
Types of Machine Learnig Algorithms(CART, ID3)Types of Machine Learnig Algorithms(CART, ID3)
Types of Machine Learnig Algorithms(CART, ID3)Fatimakhan325
 
Explicit Density Models
Explicit Density ModelsExplicit Density Models
Explicit Density ModelsSangwoo Mo
 
Build Deep Learning model to identify santader bank's dissatisfied customers
Build Deep Learning model to identify santader bank's dissatisfied customersBuild Deep Learning model to identify santader bank's dissatisfied customers
Build Deep Learning model to identify santader bank's dissatisfied customerssriram30691
 
IGARSS2011-I-Ling.ppt
IGARSS2011-I-Ling.pptIGARSS2011-I-Ling.ppt
IGARSS2011-I-Ling.pptgrssieee
 
Workshop - Introduction to Machine Learning with R
Workshop - Introduction to Machine Learning with RWorkshop - Introduction to Machine Learning with R
Workshop - Introduction to Machine Learning with RShirin Elsinghorst
 
Quantum Machine Learning for IBM AI
Quantum Machine Learning for IBM AIQuantum Machine Learning for IBM AI
Quantum Machine Learning for IBM AISasha Lazarevic
 

Similaire à ICCV2009: Max-Margin Ađitive Classifiers for Detection (20)

Data Science - Part IX - Support Vector Machine
Data Science - Part IX -  Support Vector MachineData Science - Part IX -  Support Vector Machine
Data Science - Part IX - Support Vector Machine
 
Mattar_PhD_Thesis
Mattar_PhD_ThesisMattar_PhD_Thesis
Mattar_PhD_Thesis
 
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...
[Q-tangled 22] Deconstructing Quantum Machine Learning Algorithms - Sasha Laz...
 
Guide
GuideGuide
Guide
 
Huong dan cu the svm
Huong dan cu the svmHuong dan cu the svm
Huong dan cu the svm
 
Machine learning, biomarker accuracy and best practices
Machine learning, biomarker accuracy and best practicesMachine learning, biomarker accuracy and best practices
Machine learning, biomarker accuracy and best practices
 
Optimizing Terascale Machine Learning Pipelines with Keystone ML
Optimizing Terascale Machine Learning Pipelines with Keystone MLOptimizing Terascale Machine Learning Pipelines with Keystone ML
Optimizing Terascale Machine Learning Pipelines with Keystone ML
 
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...Exploration of Supervised Machine Learning Techniques for Runtime Selection o...
Exploration of Supervised Machine Learning Techniques for Runtime Selection o...
 
SVM
SVMSVM
SVM
 
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...
Master defence 2020 -Volodymyr Lut-Neural Architecture Search: a Probabilisti...
 
4. Classification.pdf
4. Classification.pdf4. Classification.pdf
4. Classification.pdf
 
BioWeka
BioWekaBioWeka
BioWeka
 
Types of Machine Learnig Algorithms(CART, ID3)
Types of Machine Learnig Algorithms(CART, ID3)Types of Machine Learnig Algorithms(CART, ID3)
Types of Machine Learnig Algorithms(CART, ID3)
 
Explicit Density Models
Explicit Density ModelsExplicit Density Models
Explicit Density Models
 
Build Deep Learning model to identify santader bank's dissatisfied customers
Build Deep Learning model to identify santader bank's dissatisfied customersBuild Deep Learning model to identify santader bank's dissatisfied customers
Build Deep Learning model to identify santader bank's dissatisfied customers
 
Guide
GuideGuide
Guide
 
IGARSS2011-I-Ling.ppt
IGARSS2011-I-Ling.pptIGARSS2011-I-Ling.ppt
IGARSS2011-I-Ling.ppt
 
Workshop - Introduction to Machine Learning with R
Workshop - Introduction to Machine Learning with RWorkshop - Introduction to Machine Learning with R
Workshop - Introduction to Machine Learning with R
 
Quantum Machine Learning for IBM AI
Quantum Machine Learning for IBM AIQuantum Machine Learning for IBM AI
Quantum Machine Learning for IBM AI
 
Svm on cloud (presntation)
Svm on cloud  (presntation)Svm on cloud  (presntation)
Svm on cloud (presntation)
 

Plus de zukun

My lyn tutorial 2009
My lyn tutorial 2009My lyn tutorial 2009
My lyn tutorial 2009zukun
 
ETHZ CV2012: Tutorial openCV
ETHZ CV2012: Tutorial openCVETHZ CV2012: Tutorial openCV
ETHZ CV2012: Tutorial openCVzukun
 
ETHZ CV2012: Information
ETHZ CV2012: InformationETHZ CV2012: Information
ETHZ CV2012: Informationzukun
 
Siwei lyu: natural image statistics
Siwei lyu: natural image statisticsSiwei lyu: natural image statistics
Siwei lyu: natural image statisticszukun
 
Lecture9 camera calibration
Lecture9 camera calibrationLecture9 camera calibration
Lecture9 camera calibrationzukun
 
Brunelli 2008: template matching techniques in computer vision
Brunelli 2008: template matching techniques in computer visionBrunelli 2008: template matching techniques in computer vision
Brunelli 2008: template matching techniques in computer visionzukun
 
Modern features-part-4-evaluation
Modern features-part-4-evaluationModern features-part-4-evaluation
Modern features-part-4-evaluationzukun
 
Modern features-part-3-software
Modern features-part-3-softwareModern features-part-3-software
Modern features-part-3-softwarezukun
 
Modern features-part-2-descriptors
Modern features-part-2-descriptorsModern features-part-2-descriptors
Modern features-part-2-descriptorszukun
 
Modern features-part-1-detectors
Modern features-part-1-detectorsModern features-part-1-detectors
Modern features-part-1-detectorszukun
 
Modern features-part-0-intro
Modern features-part-0-introModern features-part-0-intro
Modern features-part-0-introzukun
 
Lecture 02 internet video search
Lecture 02 internet video searchLecture 02 internet video search
Lecture 02 internet video searchzukun
 
Lecture 01 internet video search
Lecture 01 internet video searchLecture 01 internet video search
Lecture 01 internet video searchzukun
 
Lecture 03 internet video search
Lecture 03 internet video searchLecture 03 internet video search
Lecture 03 internet video searchzukun
 
Icml2012 tutorial representation_learning
Icml2012 tutorial representation_learningIcml2012 tutorial representation_learning
Icml2012 tutorial representation_learningzukun
 
Advances in discrete energy minimisation for computer vision
Advances in discrete energy minimisation for computer visionAdvances in discrete energy minimisation for computer vision
Advances in discrete energy minimisation for computer visionzukun
 
Gephi tutorial: quick start
Gephi tutorial: quick startGephi tutorial: quick start
Gephi tutorial: quick startzukun
 
EM algorithm and its application in probabilistic latent semantic analysis
EM algorithm and its application in probabilistic latent semantic analysisEM algorithm and its application in probabilistic latent semantic analysis
EM algorithm and its application in probabilistic latent semantic analysiszukun
 
Object recognition with pictorial structures
Object recognition with pictorial structuresObject recognition with pictorial structures
Object recognition with pictorial structureszukun
 
Iccv2011 learning spatiotemporal graphs of human activities
Iccv2011 learning spatiotemporal graphs of human activities Iccv2011 learning spatiotemporal graphs of human activities
Iccv2011 learning spatiotemporal graphs of human activities zukun
 

Plus de zukun (20)

My lyn tutorial 2009
My lyn tutorial 2009My lyn tutorial 2009
My lyn tutorial 2009
 
ETHZ CV2012: Tutorial openCV
ETHZ CV2012: Tutorial openCVETHZ CV2012: Tutorial openCV
ETHZ CV2012: Tutorial openCV
 
ETHZ CV2012: Information
ETHZ CV2012: InformationETHZ CV2012: Information
ETHZ CV2012: Information
 
Siwei lyu: natural image statistics
Siwei lyu: natural image statisticsSiwei lyu: natural image statistics
Siwei lyu: natural image statistics
 
Lecture9 camera calibration
Lecture9 camera calibrationLecture9 camera calibration
Lecture9 camera calibration
 
Brunelli 2008: template matching techniques in computer vision
Brunelli 2008: template matching techniques in computer visionBrunelli 2008: template matching techniques in computer vision
Brunelli 2008: template matching techniques in computer vision
 
Modern features-part-4-evaluation
Modern features-part-4-evaluationModern features-part-4-evaluation
Modern features-part-4-evaluation
 
Modern features-part-3-software
Modern features-part-3-softwareModern features-part-3-software
Modern features-part-3-software
 
Modern features-part-2-descriptors
Modern features-part-2-descriptorsModern features-part-2-descriptors
Modern features-part-2-descriptors
 
Modern features-part-1-detectors
Modern features-part-1-detectorsModern features-part-1-detectors
Modern features-part-1-detectors
 
Modern features-part-0-intro
Modern features-part-0-introModern features-part-0-intro
Modern features-part-0-intro
 
Lecture 02 internet video search
Lecture 02 internet video searchLecture 02 internet video search
Lecture 02 internet video search
 
Lecture 01 internet video search
Lecture 01 internet video searchLecture 01 internet video search
Lecture 01 internet video search
 
Lecture 03 internet video search
Lecture 03 internet video searchLecture 03 internet video search
Lecture 03 internet video search
 
Icml2012 tutorial representation_learning
Icml2012 tutorial representation_learningIcml2012 tutorial representation_learning
Icml2012 tutorial representation_learning
 
Advances in discrete energy minimisation for computer vision
Advances in discrete energy minimisation for computer visionAdvances in discrete energy minimisation for computer vision
Advances in discrete energy minimisation for computer vision
 
Gephi tutorial: quick start
Gephi tutorial: quick startGephi tutorial: quick start
Gephi tutorial: quick start
 
EM algorithm and its application in probabilistic latent semantic analysis
EM algorithm and its application in probabilistic latent semantic analysisEM algorithm and its application in probabilistic latent semantic analysis
EM algorithm and its application in probabilistic latent semantic analysis
 
Object recognition with pictorial structures
Object recognition with pictorial structuresObject recognition with pictorial structures
Object recognition with pictorial structures
 
Iccv2011 learning spatiotemporal graphs of human activities
Iccv2011 learning spatiotemporal graphs of human activities Iccv2011 learning spatiotemporal graphs of human activities
Iccv2011 learning spatiotemporal graphs of human activities
 

Dernier

On National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan FellowsOn National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan FellowsMebane Rash
 
1029 - Danh muc Sach Giao Khoa 10 . pdf
1029 -  Danh muc Sach Giao Khoa 10 . pdf1029 -  Danh muc Sach Giao Khoa 10 . pdf
1029 - Danh muc Sach Giao Khoa 10 . pdfQucHHunhnh
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introductionMaksud Ahmed
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfciinovamais
 
Spellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please PractiseSpellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please PractiseAnaAcapella
 
This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.christianmathematics
 
Dyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxDyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxcallscotland1987
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxAreebaZafar22
 
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdfUGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdfNirmal Dwivedi
 
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in DelhiRussian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhikauryashika82
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfagholdier
 
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...ZurliaSoop
 
SOC 101 Demonstration of Learning Presentation
SOC 101 Demonstration of Learning PresentationSOC 101 Demonstration of Learning Presentation
SOC 101 Demonstration of Learning Presentationcamerronhm
 
Unit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptxUnit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptxVishalSingh1417
 
ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701bronxfugly43
 
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...pradhanghanshyam7136
 
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxSKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxAmanpreet Kaur
 
Seal of Good Local Governance (SGLG) 2024Final.pptx
Seal of Good Local Governance (SGLG) 2024Final.pptxSeal of Good Local Governance (SGLG) 2024Final.pptx
Seal of Good Local Governance (SGLG) 2024Final.pptxnegromaestrong
 
Application orientated numerical on hev.ppt
Application orientated numerical on hev.pptApplication orientated numerical on hev.ppt
Application orientated numerical on hev.pptRamjanShidvankar
 

Dernier (20)

On National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan FellowsOn National Teacher Day, meet the 2024-25 Kenan Fellows
On National Teacher Day, meet the 2024-25 Kenan Fellows
 
1029 - Danh muc Sach Giao Khoa 10 . pdf
1029 -  Danh muc Sach Giao Khoa 10 . pdf1029 -  Danh muc Sach Giao Khoa 10 . pdf
1029 - Danh muc Sach Giao Khoa 10 . pdf
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introduction
 
Spatium Project Simulation student brief
Spatium Project Simulation student briefSpatium Project Simulation student brief
Spatium Project Simulation student brief
 
Activity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdfActivity 01 - Artificial Culture (1).pdf
Activity 01 - Artificial Culture (1).pdf
 
Spellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please PractiseSpellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please Practise
 
This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.
 
Dyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxDyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptx
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptx
 
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdfUGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
UGC NET Paper 1 Mathematical Reasoning & Aptitude.pdf
 
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in DelhiRussian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
Russian Escort Service in Delhi 11k Hotel Foreigner Russian Call Girls in Delhi
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdf
 
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
Jual Obat Aborsi Hongkong ( Asli No.1 ) 085657271886 Obat Penggugur Kandungan...
 
SOC 101 Demonstration of Learning Presentation
SOC 101 Demonstration of Learning PresentationSOC 101 Demonstration of Learning Presentation
SOC 101 Demonstration of Learning Presentation
 
Unit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptxUnit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptx
 
ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701
 
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
 
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxSKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
 
Seal of Good Local Governance (SGLG) 2024Final.pptx
Seal of Good Local Governance (SGLG) 2024Final.pptxSeal of Good Local Governance (SGLG) 2024Final.pptx
Seal of Good Local Governance (SGLG) 2024Final.pptx
 
Application orientated numerical on hev.ppt
Application orientated numerical on hev.pptApplication orientated numerical on hev.ppt
Application orientated numerical on hev.ppt
 

ICCV2009: Max-Margin Ađitive Classifiers for Detection

  • 1. Max-Margin Additive Classifiers for Detection SubhransuMaji & Alexander Berg University of California at Berkeley Columbia University ICCV 2009, Kyoto, Japan
  • 2. Accuracy vs. Evaluation Timefor SVM Classifiers Non-linear Kernel Evaluation time Linear Kernel Accuracy
  • 3. Accuracy vs. Evaluation Timefor SVM Classifiers Non-linear Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy
  • 4. Non-linear Kernel Additive Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy Accuracy vs. Evaluation Timefor SVM Classifiers
  • 5. Additive Kernel Non-linear Kernel Additive Kernel Evaluation time Our CVPR 08 Linear Kernel Accuracy Accuracy vs. Evaluation Timefor SVM Classifiers
  • 6. Accuracy vs. Evaluation Timefor SVM Classifiers Additive Kernel Non-linear Kernel Evaluation time Our CVPR 08 Linear Kernel Additive Kernel Accuracy Made it possible to use SVMs with additive kernels for detection.
  • 7. Additive Classifiers Much work already uses them! SVMs with additive kernels are additive classifiers Histogram based kernels Histogram intersection, chi-squared kernel Pyramid Match Kernel (Grauman & Darell, ICCV’05) Spatial Pyramid Match Kernel (Lazebniket.al., CVPR’06) ….
  • 8. Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time Linear Kernel Accuracy
  • 9. Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time <=1990s Linear Accuracy
  • 10. Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time Today Linear Accuracy Eg. Cutting Plane, Stoc. Gradient Descend, Dual Coordinate Descend
  • 11. Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time Our CVPR 08 Linear Accuracy
  • 12. Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time Our CVPR 08 ✗ Linear Accuracy
  • 13. Accuracy vs. Training Timefor SVM Classifiers Non-linear Additive Training time This Paper Linear Accuracy
  • 14. Accuracy vs. Training Timefor SVM Classifiers Non-linear Training time This Paper Linear Additive Accuracy Makes it possible to train additive classifiers very fast.
  • 15. Summary Additive classifiers are widely used and can provide better accuracy than linear Our CVPR 08: SVMs with additive kernels are additive classifiers and can be evaluated in O(#dim) -- same as linear. This work: additive classifiers can be trained directly as efficiently (up to a small constant) as the best approaches for training linear classifiers. An example
  • 16.
  • 17. Can learn non-linear boundaries in input space Classification Function Kernel Trick
  • 18. Embeddings… These embeddings can be high dimensional (even infinite) Our approach is based on embeddings thatapproximate kernels. We’d like this to be as accurate as possible We are going to use fast linear classifier training algorithms on the so sparseness is important.
  • 19. Key Idea: Embedding an Additive Kernel Additive Kernels are easy to embed, just embed each dimension independently Linear Embedding for min Kernel for integers For non integers can approximate by quantizing
  • 20. Issues: Embedding Error Quantization leads to large errors Better encoding x y
  • 21. Issues: Sparsity Represent with sparse values
  • 22. Linear SVM objective (solve with LIBLINEAR): Encoded SVM objective (not practical): Linear vs. Encoded SVMs
  • 23. Linear vs. Encoded SVMs Linear SVM objective (solve with LIBLINEAR): Encoded SVM modified (custom solver): Encourages smooth functions Closely approximates min kernel SVM Custom solver : PWLSGD (see paper)
  • 24. Linear SVM objective (solve with LIBLINEAR): Encoded SVM objective (solve with LIBLINEAR) : Linear vs. Encoded SVMs
  • 25. Additive Classifier Choices Regularization Encoding
  • 26. Additive Classifier Choices Accuracy Increases Regularization Encoding Evaluation times are similar
  • 27. Additive Classifier Choices Accuracy Increases Regularization Encoding Accuracy Increases Evaluation times are similar
  • 28. Additive Classifier Choices Accuracy Increases Regularization Encoding Accuracy Increases Standard solver Eg. LIBSVM Few lines of code + standard solver Eg. LIBLINEAR
  • 29. Additive Classifier Choices Accuracy Increases Regularization Encoding Accuracy Increases Custom solver
  • 30. Additive Classifier Choices Accuracy Increases Regularization Encoding Accuracy Increases Classifier Notations
  • 31. Experiments “Small” Scale: Caltech 101 (Fei-Fei, et.al.) “Medium” Scale: DC Pedestrians (Munder & Gavrila) “Large” Scale : INRIA Pedestrians (Dalal & Triggs)
  • 32. Experiment : DC Pedestrians (3.18s, 89.25%) (1.86s, 88.80%) (363s, 89.05%) (2.98s, 85.71%) 100x faster training time ~ linear SVM accuracy ~ kernel SVM (1.89s, 72.98%) 20,000 features, 656 dimensional 100 bins for encoding 6-fold cross validation
  • 33. Experiment : Caltech 101 (291s, 55.35%) (2687s, 56.49%) (102s, 54.8%) (90s, 51.64%) 10x faster Small loss in accuracy (41s, 46.15%) 30 training examples per category 100 bins for encoding Pyramid HOG + Spatial Pyramid Match Kernel
  • 34. Experiment : INRIA Pedestrians (140 mins, 0.95) (76s, 0.94) (27s, 0.88) 300x faster training time ~ linear SVM accuracy ~ kernel SVMtrains the detector in < 2 mins (122s, 0.85) (20s, 0.82) SPHOG: 39,000 features, 2268 dimensional 100 bins for encoding Cross Validation Plots
  • 35. Experiment : INRIA Pedestrians 300x faster training time ~ linear SVM accuracy ~ kernel SVMtrains the detector in < 2 mins SPHOG: 39,000 features, 2268 dimensional 100 bins for encoding Cross Validation Plots
  • 36. Take Home Messages Additive models are practical for large scale data Can be trained discriminatively: Poor man’s version : encode + Linear SVM Solver Middle man’s version : encode + Custom Solver Rich man’s version : Min Kernel SVM Embedding only Approximates kernels, leads to small loss in accuracy but up to 100x speedup in training time Everyone should use: see code on our websites Fast IKSVM from CVPR’08, Encoded SVMs, etc

Notes de l'éditeur

  1. Thankyou. Good afternoon everybody. I am going to present ways to train additive classifiers efficiently . This work is a part of an ongoing collaboration with alex berg.
  2. For any classification task the two main things we care about are accuracy and evaluation time. Especially for object detection where one evalutaes a classifier on thousands of windowsPer image – the evalutation time becomes very important. In the past linear SVMs though relatively less accurate were preferred over kernel SVMs for real-time applications.
  3. In our CVPR 08 paper…
  4. We identified a subset of non-linear kernels, called additive kernels that are used in many of the current object recognition tasks. These kernels have the special form that they decompose as a sum of Kernels over individual dimensions.
  5. We identified a subset of non-linear kernels, called additive kernels that are used in many of the current object recognition tasks. These kernels have the special form that they decompose as a sum of Kernels over individual dimensions.
  6. And showed that they can be evaulated efficiently. This makes it possible for one to use more accurate classifiers with relatively no loss in speed. In fact more than half of thisYear’s submissions to the PACCAL VOC object detection challenge use variants of additive kernels.
  7. In this talk we are going to talk about additive models in general – where the classifier decomposes into dimensions. This may seem restrictive but it’s a useful class of classifiers which iis strictly more general than linear classifiers.In fact if the underlying kernel for the SVM is additive then the classifier is also additive
  8. Pic looks similar to that for evaluation time… it is important to note that this was not the case even somewhat recently…
  9. Maybe put some refs on this…
  10. Maybe put some refs on this…As mentioned before, our previous work identified a subset of non-linear classifiers with an additive structure and showed they could be evaluated efficiently, but unfortunately did not address improving efficiency for training…
  11. Maybe put some refs on this…
  12. This paper addresses efficient training for additive classifiers, developing training methods that are about as efficient as the best methods fortraining linear classifiers. We also demonstrate the accuracy avantages on some popular datasets.?....
  13. Should we change the wording? Drop SVM?
  14. (finish this by 5 mins)
  15. The idea of support vector machines is to find a separating hyperplane on the data into a high dimension space using a Kernel.The final classifier is ofcouse a line in a very high dimensional space but can be expressed using only the Kernel function using the so called kernel trick. If the embedded space is low dimensional then one can take advantage of the very fast linear SVM training algorithms which scale linearly with trainingData as opposed to the quadratic growth for the kernel SVM.
  16. Unfortunately these embeddings are often high dimensionalOur approach can be seen as finding embeddings that are both sparse and accurate so that we can use the very best of the linear SVM training algorithms for trainingThe classifier. In fact we would ideally like the number of non zero entries in the embedded features to be a small multiple of the nonn zero entries in the input features.
  17. A key idea of the paper is to realize that additive kernels are easy to embed as the final embedding is just a concatenation of the individual dimension embeddingsAS as example the min kernel or the histogram intersection kernel defined as A well known embedding for min kernel for integers is the unaryencoding where each number is represented in the unaryExample …For non-integers one may just approximate this by quantization