SlideShare ist ein Scribd-Unternehmen logo
1 von 45
Stanford-Nokia CollaborationMobile Augmented RealityAugust 2009 Review,[object Object],	Bernd Girod		Radek Grzeszczuk,[object Object],              	Stanford University	Nokia Research Center,[object Object]
Mobile Augmented Reality Team,[object Object],Radek Grzeszczuk,[object Object],Bernd Girod,[object Object],Vijay Chandrasekhar,[object Object],Gabriel Takacs,[object Object],Wei-Chao Chen,[object Object],Natasha Gelfand,[object Object],Yingen Xiong,[object Object],Kari Pulli,[object Object],Sam Tsai,[object Object],David Chen,[object Object],Jana Kosecka,[object Object],Ramakrishna Vedantham,[object Object],Mina Makar,[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-viewvocabulary trees,[object Object],Matching against 3-d models,[object Object],Summary and future directions,[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-viewvocabulary trees,[object Object],Matching against 3-d models,[object Object],Summary and future directions,[object Object]
Mobile Visual Search,[object Object],User takes picture,[object Object],… chooses action          …,[object Object],…confirms POI    ,[object Object]
Mobile Visual Search Applications ,[object Object],Museum Guide,[object Object],Tourist Guide,[object Object],Landmarks,[object Object],Wine Labels,[object Object],Comparison Shopping,[object Object],Ads/Catalogs,[object Object],CDs/DVDs/Books,[object Object],Movie Posters,[object Object]
GPS,[object Object],Server,[object Object],Landmark Recognition withFeature Matching on the Phone,[object Object],Memorial Church,[object Object]
Prefetched Data,[object Object],“Bag of Words” Matching,[object Object],Query Image,[object Object],Geometric,[object Object],Consistency,[object Object],Check,[object Object],Feature,[object Object],Descriptors,[object Object],Feature,[object Object],Correspondences,[object Object],Database Images,[object Object]
Computing Visual Words,[object Object],dx,[object Object],dy,[object Object],scale,[object Object],SIFT Descriptor,[object Object],SURF Descriptor,[object Object],y,[object Object],x,[object Object],Σdx,[object Object],Σdy,[object Object],Σ|dx|,[object Object],Σ|dy|,[object Object],Σ,[object Object],Σ,[object Object],Σ,[object Object],Σ,[object Object],Σ,[object Object],Σ,[object Object],Σ,[object Object],Σ    ,[object Object],Color,[object Object],Gray,[object Object],Dxx,[object Object],Σdx,[object Object],Σdy,[object Object],Σ|dx|,[object Object],Σ|dy|,[object Object],Maxima,[object Object],Dxy,[object Object],…,[object Object],…,[object Object],DxxDyy-(0.9Dxy)2,[object Object],Σdx,[object Object],Σdy,[object Object],Σ|dx|,[object Object],Σ|dy|,[object Object],Dyy,[object Object],Orient along ,[object Object],dominant gradient,[object Object],Oriented Patch,[object Object],Gradient Field,[object Object],Filters,[object Object],Blob Response,[object Object]
Matching Performance,[object Object],~90 images/kernel,[object Object],~90 images/kernel,[object Object],~1000 images/kernel,[object Object],True Matches,[object Object],False Matches,[object Object]
Timing Analysis(Q2 2008),[object Object],Nokia N95,[object Object],332 MHz ARM,[object Object],64 MB RAM ,[object Object],100 KByte JPEG; uplink 60 Kbps,[object Object],Downloads,[object Object],Upload,[object Object],Upload,[object Object],Geometric,[object Object],Consistency,[object Object],Extract,[object Object],Features,[object Object],Extract,[object Object],Features,[object Object],Feature Matching,[object Object],Extract Features ,[object Object],on Phone,[object Object],All on Phone,[object Object],All on Server,[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-viewvocabulary trees,[object Object],Matching against 3-d models,[object Object],Summary and future directions,[object Object]
Advanced Feature Compression,[object Object],Transform Coding of SIFT/SURF descriptors[Chandrasekhar et al.,  VCIP 09],[object Object],Direct compression of oriented image patch [M. Makar et al., ICASSP 09],[object Object],Descriptor designed for compressibility: CHoG[Chandrasekhar et al.,  CVPR 09],[object Object],Tree-Structured Vector QuantizationTree Histogram Coding [Chen et al.,  DCC 09],[object Object],Compression of Location Information[Tsai et al.,  Mobimedia 09],[object Object]
Patch,[object Object],CHoG: Compressed Histogram of Gradients,[object Object],Gradient distributions,[object Object],for each bin,[object Object],Gradients,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dx,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],dy,[object Object],Spatial,[object Object],binning,[object Object],01101,[object Object],101101,[object Object],Histogram,[object Object],compression,[object Object],0100011,[object Object],111001,[object Object],0010011,[object Object],01100,[object Object],1010100,[object Object],CHoGDescriptor,[object Object]
CHoG: Histogram Compression,[object Object],0.46,[object Object],1/2,[object Object],0.21,[object Object],1/4,[object Object],0.46,[object Object],0.16,[object Object],1/8,[object Object],  0.09,[object Object],  0.08,[object Object],1/16,[object Object],1/16,[object Object],0.21,[object Object],Gradient distribution,[object Object],0.08,[object Object],0.16,[object Object],0.09,[object Object],Huffman treeapproximatesprobabilities,[object Object],Gradient binning,[object Object]
Enumerating Huffman Trees,[object Object],Rooted binary trees with nleaf nodes,[object Object]
Feature Matching Performance,[object Object],Tree Structured Vector Quantizer,[object Object],SURF Transform,[object Object],Random,[object Object],Projections,[object Object],BoostSSC,[object Object],Patch + SIFT,[object Object],CHoG,[object Object],SIFT Transform,[object Object],Ground truth data setof matching patches,[object Object],Descriptor Size (bits),[object Object],[Winder & Brown CVPR ’07],[object Object]
Compressed Domain Matching,[object Object],1   2   3    4   5   6 ,[object Object],1,[object Object],2,[object Object],3,[object Object],4,[object Object],5,[object Object],6 ,[object Object],Dist(·),[object Object],Distance,[object Object],Distance,[object Object],Look-up table,[object Object],Tree index,[object Object],Gradient binning,[object Object],Gradient distribution,[object Object]
Nearest Neighbor Search,[object Object],372,[object Object],Exact,[object Object],ANN0.3 % errors,[object Object],Exact,[object Object],47,[object Object],28,[object Object],400,[object Object],350,[object Object],300,[object Object],250,[object Object],Query Time (sec),[object Object],200,[object Object],150,[object Object],100,[object Object],50,[object Object],0,[object Object],SIFT,[object Object],CHoG,[object Object],106 database descriptors,[object Object],103 query descriptors,[object Object]
Location Histogram Coding,[object Object],Feature,[object Object],Locations,[object Object],(x,y),[object Object],Spatial,[object Object],Binning,[object Object],Context-based,[object Object],Arithmetic Coding,[object Object],-,[object Object],Refinement Bits,[object Object],Quantize,[object Object],+,[object Object],[Tsai et al., MobiMedia 2009],[object Object]
Compressed Feature Vector,[object Object],52,[object Object],84,[object Object],1024,[object Object],1088,[object Object],59,[object Object],Size (bits),[object Object],SIFT,[object Object],Location x,y,[object Object],1088 bits,[object Object],CHoG ,[object Object],Location x,y,[object Object],~ 84 bits,[object Object],Compressedx,yCHoG,[object Object],~ 59 bits,[object Object],[Tsai et al., MobiMedia 2009],[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-viewvocabulary trees,[object Object],Matching against 3-d models,[object Object],Summary and future directions,[object Object]
Pairwise Comparison,[object Object],“Bag of Words” Matching & Affine Consistency Check,[object Object]
Growing Vocabulary Tree,[object Object],[Nistér and Stewenius, 2006],[object Object]
Growing Vocabulary Tree,[object Object],[Nistér and Stewenius, 2006],[object Object]
Growing Vocabulary Tree,[object Object],[Nistér and Stewenius, 2006],[object Object]
Growing Vocabulary Tree,[object Object],k = 3,[object Object],[Nistér and Stewenius, 2006],[object Object]
k = 3,[object Object],Growing Vocabulary Tree,[object Object],[Nistér and Stewenius, 2006],[object Object]
Querying Vocabulary Tree,[object Object],Query,[object Object]
Recognition Accuracy,[object Object],Forestof 6 trees,[object Object],Recall (Percent),[object Object],Singlevocabulary,[object Object],tree,[object Object],Number of database images,[object Object]
Vocabulary Forest,[object Object],SVT,[object Object],Features,[object Object],…,[object Object],…,[object Object],Image,[object Object],…,[object Object],Image,[object Object],…,[object Object],IFS,[object Object],Count,[object Object],…,[object Object],Count,[object Object],…,[object Object],Early Termination,[object Object],GCC,[object Object],…,[object Object],Combine Matches,[object Object]
Real-time System: Send Image,[object Object],Image,[object Object],Wireless,[object Object],Network,[object Object],Information,[object Object],Server,[object Object],VocTreeImage ,[object Object],Matching,[object Object],Feature ,[object Object],Extraction,[object Object],Camera,[object Object],Client,[object Object]
Features,[object Object],Wireless,[object Object],Network,[object Object],Information,[object Object],Server,[object Object],VocTree,[object Object],Image,[object Object],Matching,[object Object],FeatureExtraction,[object Object],Camera,[object Object],Client,[object Object],Coding,[object Object],Real-time System: Send Features,[object Object]
Timing Analysis,[object Object],Nokia N95,[object Object],332 MHz ARM,[object Object],64 MB RAM ,[object Object],Server Delay,[object Object],Execution Time (sec),[object Object],Upload,[object Object],Image,[object Object],40 kByte,[object Object],Server Delay,[object Object],Upload Features,[object Object],2.2 kByte,[object Object],Extract Features,[object Object],“Send Features”            “Send Image”,[object Object]
Timing Analysis,[object Object],Nokia N95,[object Object],332 MHz ARM,[object Object],64 MB RAM ,[object Object],Execution Time (sec),[object Object],Server Delay,[object Object],Upload,[object Object],Image,[object Object],40 kByte,[object Object],Server Delay,[object Object],Upload 2.2 kByte,[object Object],Extract Features,[object Object],“Send Features”            “Send Image”,[object Object]
Timing Analysis,[object Object],Nokia N95,[object Object],332 MHz ARM,[object Object],64 MB RAM ,[object Object],Execution Time (sec),[object Object],Server Delay,[object Object],Server Delay,[object Object],Extract Features,[object Object],“Send Features”            “Send Image”,[object Object]
Streaming MAR,[object Object],Server,[object Object],Extract Features,[object Object],Search K-D Tree,[object Object],Check Geometry,[object Object],Send Query Frame,[object Object],Send ID and Geometry,[object Object],Network,[object Object],Low Motion,[object Object],John Mayer,[object Object],Inside Wants Out,[object Object],Display ID and Draw Boundary,[object Object],CompensateCamera Pose,[object Object],Time,[object Object],High Motion,[object Object],Client,[object Object],TrackCamera Pose,[object Object],…,[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-view vocabulary trees,[object Object],City-scale landmark recognition using view invariant matching,[object Object],Summary and future directions,[object Object]
Multiview Database,[object Object],Front View Images,[object Object],Top View Images,[object Object],Bottom View Images,[object Object],Right View Images,[object Object],Left View Images,[object Object]
Multiview Vocabulary Trees,[object Object],Left,[object Object],Front,[object Object],Top,[object Object],Bottom,[object Object],Right,[object Object],Query Image,[object Object],Select Top Matches,[object Object],Select Top Matches,[object Object],Select Top Matches,[object Object],Select Top Matches,[object Object],Select Top Matches,[object Object],Geometric Consistency Check ,[object Object],Top Match,[object Object]
Multiview Matching Performance,[object Object],Front SVT,[object Object],Multiview SVTs,[object Object],Image Recall,[object Object],Match Rate ,[object Object],Query View,[object Object],Query View,[object Object],Top,[object Object],Right,[object Object],Bottom,[object Object],Right,[object Object],Front,[object Object],Left,[object Object],Top,[object Object],Bottom,[object Object],Front,[object Object],Left,[object Object]
Compact Architectural Models from Geo-Registered Image Collections,[object Object],GPS-tagged Images,[object Object],Building Outline,[object Object],Camera Poses Estimation,[object Object],Robust Map Alignment,[object Object],Efficient View,[object Object],Selection,[object Object],3D Model of Landmark,[object Object],Unstructured Image Collections: Panoramio,[object Object],Structured Image Collections: Street View data (Navteq),[object Object],[Grzeszczuk, 3DIM 2009],[object Object]
View-Invariant Matching Pipeline,[object Object],Feature,[object Object],Store,[object Object],Feature Extraction,[object Object],Image,[object Object],Database,[object Object],Rectified,[object Object],Database Images,[object Object],Image Rectification using 3D Model,[object Object],Feature Extraction,[object Object],Matching,[object Object],Results,[object Object],Oblique,[object Object],Query Image,[object Object],Rectified,[object Object],Query Image,[object Object],Image Rectification using Vanishing Points,[object Object]
Outline,[object Object],Review: landmark recognition system,[object Object],Architecture: location-based pre-fetching and matching on the phone,[object Object],Computer vision: “Bag of Words” matching,[object Object],Feature compression for server-side matching,[object Object],Approaches explored: Transform coding of features, patch compression,[object Object],Compressible descriptor: CHoG (Compressed Histogram of Gradients),[object Object],Scalability for large data bases,[object Object],From “Bags of Words” to “Vocabulary Trees” to “Vocabulary Forests”,[object Object],Accuracy vs. data base size,[object Object],Towards 3D,[object Object],Multi-viewvocabulary trees,[object Object],Matching against 3-d models,[object Object],Summary and future directions,[object Object]
Research Directions,[object Object],Research area: image features,[object Object],Keypoint detection optimized for CHoG, prioritization,[object Object],Comprehensive performance analysis of compressed feature matching,[object Object],Next generation CHoG: soft kernels vs. hard binning, embedded, refinablebitstream,[object Object],Beyond RANSAC: advanced geometry matching and coding, incorporate scale and orientation,[object Object],Research area: image database/vocabulary trees,[object Object],Optimum tree/forest growing, CHoG trees, incremental data base update,[object Object],Fast query, early termination, distance metrics, scoring, nearest neighbor algorithms,[object Object],Trees for phone implementation, inverted file caching, tree histogram coding,[object Object],Research area: streaming mobile augmented reality,[object Object],Camera pose estimation, feature tracking, temporally coherent feature extraction,[object Object],Continuous recognition strategies, scheduling, latency minimization,[object Object],Superposition of graphics information, motion compensation, occlusion handling,[object Object],Research area: 3D modeling,[object Object],Image matching pipeline using 3D models,[object Object],Automatic image rectification, features from texture maps,[object Object],Methods for integrating heterogeneous image sources,[object Object],Demonstrate improved landmark recognition for large-scale urban scene,[object Object],Collaboration with Marc Pollefeys, ETH Zurich,[object Object]

Weitere ähnliche Inhalte

Ähnlich wie Nokia Augmented Reality

Scientific Applications and Heterogeneous Architectures
Scientific Applications and Heterogeneous ArchitecturesScientific Applications and Heterogeneous Architectures
Scientific Applications and Heterogeneous Architecturesinside-BigData.com
 
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...Big Data Spain
 
Bayesian Network 을 활용한 예측 분석
Bayesian Network 을 활용한 예측 분석Bayesian Network 을 활용한 예측 분석
Bayesian Network 을 활용한 예측 분석datasciencekorea
 
Fishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data LakeFishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data LakeArangoDB Database
 
Cogent3 d master slides (12 april 2009)
Cogent3 d master slides (12 april 2009)Cogent3 d master slides (12 april 2009)
Cogent3 d master slides (12 april 2009)Danny Bronson
 
A general framework for predicting the optimal computing configuration for cl...
A general framework for predicting the optimal computing configuration for cl...A general framework for predicting the optimal computing configuration for cl...
A general framework for predicting the optimal computing configuration for cl...Scott Farley
 
IEEE Big data 2016 Title and Abstract
IEEE Big data  2016 Title and AbstractIEEE Big data  2016 Title and Abstract
IEEE Big data 2016 Title and Abstracttsysglobalsolutions
 
Dynamics in graph analysis (PyData Carolinas 2016)
Dynamics in graph analysis (PyData Carolinas 2016)Dynamics in graph analysis (PyData Carolinas 2016)
Dynamics in graph analysis (PyData Carolinas 2016)Benjamin Bengfort
 
Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &IAEME Publication
 
Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &IAEME Publication
 
Workflow Provenance: From Modelling to Reporting
Workflow Provenance: From Modelling to ReportingWorkflow Provenance: From Modelling to Reporting
Workflow Provenance: From Modelling to ReportingRayhan Ferdous
 
Semantic Interoperability & Information Brokering in Global Information Systems
Semantic Interoperability & Information Brokering in Global Information SystemsSemantic Interoperability & Information Brokering in Global Information Systems
Semantic Interoperability & Information Brokering in Global Information SystemsAmit Sheth
 
Coping with Data Variety in the Big Data Era: The Semantic Computing Approach
Coping with Data Variety in the Big Data Era: The Semantic Computing ApproachCoping with Data Variety in the Big Data Era: The Semantic Computing Approach
Coping with Data Variety in the Big Data Era: The Semantic Computing ApproachAndre Freitas
 
V like Velocity, Predicting in Real-Time with Azure ML
V like Velocity, Predicting in Real-Time with Azure MLV like Velocity, Predicting in Real-Time with Azure ML
V like Velocity, Predicting in Real-Time with Azure MLBarbara Fusinska
 
The interoperability challenges of 3D personal data
The interoperability challenges of 3D personal dataThe interoperability challenges of 3D personal data
The interoperability challenges of 3D personal dataJuan V. Dura
 

Ähnlich wie Nokia Augmented Reality (20)

Scientific Applications and Heterogeneous Architectures
Scientific Applications and Heterogeneous ArchitecturesScientific Applications and Heterogeneous Architectures
Scientific Applications and Heterogeneous Architectures
 
Fishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data Lake Fishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data Lake
 
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...
Fishing Graphs in a Hadoop Data Lake by Jörg Schad and Max Neunhoeffer at Big...
 
Dagstuhl 2013 - Montali - On the Relationship between OBDA and Relational Map...
Dagstuhl 2013 - Montali - On the Relationship between OBDA and Relational Map...Dagstuhl 2013 - Montali - On the Relationship between OBDA and Relational Map...
Dagstuhl 2013 - Montali - On the Relationship between OBDA and Relational Map...
 
3DRepo
3DRepo3DRepo
3DRepo
 
Bayesian Network 을 활용한 예측 분석
Bayesian Network 을 활용한 예측 분석Bayesian Network 을 활용한 예측 분석
Bayesian Network 을 활용한 예측 분석
 
Fishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data LakeFishing Graphs in a Hadoop Data Lake
Fishing Graphs in a Hadoop Data Lake
 
Cogent3 d master slides (12 april 2009)
Cogent3 d master slides (12 april 2009)Cogent3 d master slides (12 april 2009)
Cogent3 d master slides (12 april 2009)
 
A general framework for predicting the optimal computing configuration for cl...
A general framework for predicting the optimal computing configuration for cl...A general framework for predicting the optimal computing configuration for cl...
A general framework for predicting the optimal computing configuration for cl...
 
BigData
BigDataBigData
BigData
 
IEEE Big data 2016 Title and Abstract
IEEE Big data  2016 Title and AbstractIEEE Big data  2016 Title and Abstract
IEEE Big data 2016 Title and Abstract
 
Dynamics in graph analysis (PyData Carolinas 2016)
Dynamics in graph analysis (PyData Carolinas 2016)Dynamics in graph analysis (PyData Carolinas 2016)
Dynamics in graph analysis (PyData Carolinas 2016)
 
Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &
 
Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &Developing and comparing an encoding system using vector quantization &
Developing and comparing an encoding system using vector quantization &
 
Workflow Provenance: From Modelling to Reporting
Workflow Provenance: From Modelling to ReportingWorkflow Provenance: From Modelling to Reporting
Workflow Provenance: From Modelling to Reporting
 
Semantic Interoperability & Information Brokering in Global Information Systems
Semantic Interoperability & Information Brokering in Global Information SystemsSemantic Interoperability & Information Brokering in Global Information Systems
Semantic Interoperability & Information Brokering in Global Information Systems
 
Coping with Data Variety in the Big Data Era: The Semantic Computing Approach
Coping with Data Variety in the Big Data Era: The Semantic Computing ApproachCoping with Data Variety in the Big Data Era: The Semantic Computing Approach
Coping with Data Variety in the Big Data Era: The Semantic Computing Approach
 
V like Velocity, Predicting in Real-Time with Azure ML
V like Velocity, Predicting in Real-Time with Azure MLV like Velocity, Predicting in Real-Time with Azure ML
V like Velocity, Predicting in Real-Time with Azure ML
 
G1803054653
G1803054653G1803054653
G1803054653
 
The interoperability challenges of 3D personal data
The interoperability challenges of 3D personal dataThe interoperability challenges of 3D personal data
The interoperability challenges of 3D personal data
 

Kürzlich hochgeladen

Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioChristian Posta
 
Secure your environment with UiPath and CyberArk technologies - Session 1
Secure your environment with UiPath and CyberArk technologies - Session 1Secure your environment with UiPath and CyberArk technologies - Session 1
Secure your environment with UiPath and CyberArk technologies - Session 1DianaGray10
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostMatt Ray
 
UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7DianaGray10
 
UiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPathCommunity
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IES VE
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...DianaGray10
 
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Will Schroeder
 
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1DianaGray10
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...Aggregage
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationIES VE
 
UiPath Studio Web workshop series - Day 8
UiPath Studio Web workshop series - Day 8UiPath Studio Web workshop series - Day 8
UiPath Studio Web workshop series - Day 8DianaGray10
 
9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding TeamAdam Moalla
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?IES VE
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesDavid Newbury
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAshyamraj55
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UbiTrack UK
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdfPedro Manuel
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6DianaGray10
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxUdaiappa Ramachandran
 

Kürzlich hochgeladen (20)

Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and Istio
 
Secure your environment with UiPath and CyberArk technologies - Session 1
Secure your environment with UiPath and CyberArk technologies - Session 1Secure your environment with UiPath and CyberArk technologies - Session 1
Secure your environment with UiPath and CyberArk technologies - Session 1
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
 
UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7
 
UiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation Developers
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
 
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
Apres-Cyber - The Data Dilemma: Bridging Offensive Operations and Machine Lea...
 
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
 
UiPath Studio Web workshop series - Day 8
UiPath Studio Web workshop series - Day 8UiPath Studio Web workshop series - Day 8
UiPath Studio Web workshop series - Day 8
 
9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team
 
How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?How Accurate are Carbon Emissions Projections?
How Accurate are Carbon Emissions Projections?
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond Ontologies
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdf
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptx
 

Nokia Augmented Reality

  • 1.
  • 2.
  • 3.
  • 4.
  • 5.
  • 6.
  • 7.
  • 8.
  • 9.
  • 10.
  • 11.
  • 12.
  • 13.
  • 14.
  • 15.
  • 16.
  • 17.
  • 18.
  • 19.
  • 20.
  • 21.
  • 22.
  • 23.
  • 24.
  • 25.
  • 26.
  • 27.
  • 28.
  • 29.
  • 30.
  • 31.
  • 32.
  • 33.
  • 34.
  • 35.
  • 36.
  • 37.
  • 38.
  • 39.
  • 40.
  • 41.
  • 42.
  • 43.
  • 44.
  • 45.

Hinweis der Redaktion

  1. Only a limited number of different Huffman trees.Catalan number yields number of rooted binary trees (ordered leaves, no cross-overs)Count unique permutations
  2. Winder, Brown (Microsoft Resarch), “Learning Local Image Descriptors,” 64x64 patches. touristphotographs of the Trevi Fountain and of Yosemite Valley (920 images), and a test set consisting of images ofNotre Dame (500 images). BoostSSC –Boosting Similarity Sensitive CodingG. Shakhnarovich, P. Viola, and T. Darrell. Fast pose estimation with parameter sensitive hashing. In Proc. ICCV, 2003.Torralba et al., Small Codes and Large Image Databases for Recognition, CVPR2009.Random Projections - P. A. ChuohaoYeo and K. Ramchandran, “Rate-EfficientVisual Correspondences Using Random Projections,” 2008.
  3. Most retrieval application require NN search in some formThe descriptors for both SIFT and CHoG were computed from the sameset of patches. VQ-5 bin configuration, GLOH-9 cell configurationsand Huffman Tree Coding are used for CHoG, resulting in a45 dimensional descriptor. We observe that exact nearest neighborsearching is 10X faster for CHoG. Furthermore, CHoG is still 2Xfaster than using SIFT with ANN eps = 1, which incurs a small errorrate of 0.30%. The speed up results from the lower dimensionalityof the CHoG descriptor, and the use of look up tables for fastdistance computation.
  4. The scalable vocabulary tree is the data structure at the center of our recognition system. To construct an SVT, first we take every database CD cover and extract robust local features. These features can be SIFT, SURF, or your own favorite type. Then, all the feature descriptors from all the images are represented as vectors in a high-dimensional space. Here, they are shown as 2-dimensional vectors, but in reality, they can be 64-dimensional or 128-dimensional vectors.
  5. The scalable vocabulary tree is the data structure at the center of our recognition system. To construct an SVT, first we take every database CD cover and extract robust local features. These features can be SIFT, SURF, or your own favorite type. Then, all the feature descriptors from all the images are represented as vectors in a high-dimensional space. Here, they are shown as 2-dimensional vectors, but in reality, they can be 64-dimensional or 128-dimensional vectors.
  6. The scalable vocabulary tree is the data structure at the center of our recognition system. To construct an SVT, first we take every database CD cover and extract robust local features. These features can be SIFT, SURF, or your own favorite type. Then, all the feature descriptors from all the images are represented as vectors in a high-dimensional space. Here, they are shown as 2-dimensional vectors, but in reality, they can be 64-dimensional or 128-dimensional vectors.
  7. To impose some structure on this space, we perform hierarchical k-means clustering, the first step of which is dividing the space into k clusters using regular k-means.
  8. And then again, recursively splitting each large cluster into k smaller clusters. We repeat this process until the clusters become sufficiently small.What results from the hierarchical k-means algorithm is a tree structure, where tree nodes are the cluster centroids and their children are the subcluster centroids.
  9. Here is the same tree as on the previous slide, except the tree structure is more apparent. Once we have constructed an SVT on a server, how to process an incoming query is straightforward. For every query descriptor, we classify it by traversing the SVT greedily from top to bottom. Suppose the first descriptor follows this nearest neighbor path. The SVT knows which database images have features associated with every node, so it votes for the two images found on this path. Both the blue nodes and green nodes vote, but since the blue nodes are more discriminative, their vote counts for more. Then, another query descriptor goes down a different path and votes for other images. And so on, until all the query descriptors are classified. The final vote tally is a histogram indicating how likely each database image is a match.We notice that when both the query and database images are fronto-parallel, the voting scheme works well and will select the correct database match. This is because similar features are extracted from the query image and the matching database image, leading to their descriptors visiting many of the same nodes in the SVT.
  10. Performance drops with single tree, since nodes become less discriminative – fewer features are unique to a particular database image
  11. Feature extraction is robust against rotation and scale change. NOT robust against foreshortening.Overcome by putting multiple examples into data base that show object from different angles.
  12. One could put all these views into one vocabulary tree.Distributing views across parallel trees prevent competition among the among the features belonging to different views of the same object. Views compete only, once all the features are considered. Select the 25 top matches for each SVT based on bin count similarity, then find match with best geometric consistency.The multiview SVT approach is attractive for multi-core server, the search process through the different trees can be run in parallel
  13. ICCV: Sept/Oct Kyoto
  14. Reduce Database SizeIncrease Robustness