INTRODUCTION Heesoo Myeong, Ju Yong Chang, and Kyoung Mu Lee Department of EECS, ASRI, Seoul National University, Seoul, Korea Learning.

Slides:



Advertisements
Similar presentations
Semantic Contours from Inverse Detectors Bharath Hariharan et.al. (ICCV-11)
Advertisements

HOPS: Efficient Region Labeling using Higher Order Proxy Neighborhoods Albert Y. C. Chen 1, Jason J. Corso 1, and Le Wang 2 1 Dept. of Computer Science.
Weakly supervised learning of MRF models for image region labeling Jakob Verbeek LEAR team, INRIA Rhône-Alpes.
Automatic Image Annotation Using Group Sparsity
Multi-label Relational Neighbor Classification using Social Context Features Xi Wang and Gita Sukthankar Department of EECS University of Central Florida.
Active Learning for Streaming Networked Data Zhilin Yang, Jie Tang, Yutao Zhang Computer Science Department, Tsinghua University.
Foreground Focus: Finding Meaningful Features in Unlabeled Images Yong Jae Lee and Kristen Grauman University of Texas at Austin.
Challenges to image parsing researchers Lana Lazebnik UNC Chapel Hill sky sidewalk building road car person car mountain.
Carolina Galleguillos, Brian McFee, Serge Belongie, Gert Lanckriet Computer Science and Engineering Department Electrical and Computer Engineering Department.
LARGE-SCALE IMAGE PARSING Joseph Tighe and Svetlana Lazebnik University of North Carolina at Chapel Hill road building car sky.
Query Specific Fusion for Image Retrieval
Patch to the Future: Unsupervised Visual Prediction
Intelligent Systems Lab. Recognizing Human actions from Still Images with Latent Poses Authors: Weilong Yang, Yang Wang, and Greg Mori Simon Fraser University,
Object-centric spatial pooling for image classification Olga Russakovsky, Yuanqing Lin, Kai Yu, Li Fei-Fei ECCV 2012.
Boundary Preserving Dense Local Regions
Data Visualization STAT 890, STAT 442, CM 462
São Paulo Advanced School of Computing (SP-ASC’10). São Paulo, Brazil, July 12-17, 2010 Looking at People Using Partial Least Squares William Robson Schwartz.
Relevance Feedback Content-Based Image Retrieval Using Query Distribution Estimation Based on Maximum Entropy Principle Irwin King and Zhong Jin Nov
Learning to Detect A Salient Object Reporter: 鄭綱 (3/2)
Beyond Actions: Discriminative Models for Contextual Group Activities Tian Lan School of Computing Science Simon Fraser University August 12, 2010 M.Sc.
LARGE-SCALE NONPARAMETRIC IMAGE PARSING Joseph Tighe and Svetlana Lazebnik University of North Carolina at Chapel Hill CVPR 2011Workshop on Large-Scale.
CS335 Principles of Multimedia Systems Content Based Media Retrieval Hao Jiang Computer Science Department Boston College Dec. 4, 2007.
Abstract We present a model of curvilinear grouping using piecewise linear representations of contours and a conditional random field to capture continuity.
Graph Cut based Inference with Co-occurrence Statistics Ľubor Ladický, Chris Russell, Pushmeet Kohli, Philip Torr.
Con-Text: Text Detection Using Background Connectivity for Fine-Grained Object Classification Sezer Karaoglu, Jan van Gemert, Theo Gevers 1.
What, Where & How Many? Combining Object Detectors and CRFs
Mutual Information-based Stereo Matching Combined with SIFT Descriptor in Log-chromaticity Color Space Yong Seok Heo, Kyoung Mu Lee, and Sang Uk Lee.
Marcin Marszałek, Ivan Laptev, Cordelia Schmid Computer Vision and Pattern Recognition, CVPR Actions in Context.
City University of Hong Kong 18 th Intl. Conf. Pattern Recognition Self-Validated and Spatially Coherent Clustering with NS-MRF and Graph Cuts Wei Feng.
Classifying Visual Objects Regardless of Depictive Style Qi Wu, Peter Hall Department of Computer Science University of Bath.
Why Categorize in Computer Vision ?. Why Use Categories? People love categories!
Learning Decompositional Shape Models from Examples Alex Levinshtein Cristian Sminchisescu Sven Dickinson Sven Dickinson University of Toronto.
INTRODUCTION Heesoo Myeong and Kyoung Mu Lee Department of ECE, ASRI, Seoul National University, Seoul, Korea Tensor-based High-order.
Exploiting Context Analysis for Combining Multiple Entity Resolution Systems -Ramu Bandaru Zhaoqi Chen Dmitri V.kalashnikov Sharad Mehrotra.
Putting Context into Vision Derek Hoiem September 15, 2004.
Graph-based Text Classification: Learn from Your Neighbors Ralitsa Angelova , Gerhard Weikum : Max Planck Institute for Informatics Stuhlsatzenhausweg.
Towards Semantic Embedding in Visual Vocabulary Towards Semantic Embedding in Visual Vocabulary The Twenty-Third IEEE Conference on Computer Vision and.
The 18th Meeting on Image Recognition and Understanding 2015/7/29 Depth Image Enhancement Using Local Tangent Plane Approximations Kiyoshi MatsuoYoshimitsu.
Tell Me What You See and I will Show You Where It Is Jia Xu 1 Alexander G. Schwing 2 Raquel Urtasun 2,3 1 University of Wisconsin-Madison 2 University.
Saliency Aggregation: A Data- driven Approach Long Mai Yuzhen Niu Feng Liu Department of Computer Science, Portland State University Portland, OR,
A New Method for Automatic Clothing Tagging Utilizing Image-Click-Ads Introduction Conclusion Can We Do Better to Reduce Workload?
Effective Automatic Image Annotation Via A Coherent Language Model and Active Learning Rong Jin, Joyce Y. Chai Michigan State University Luo Si Carnegie.
Shiliang Zhang1, Qi Tian2, Gang Hua3, Qingming Huang4, Shipeng Li2 1Key Lab of Intelli. Info. Process., Inst. of Comput. Tech., CAS, Beijing , China.
Recognition Using Visual Phrases
Object-Graphs for Context-Aware Category Discovery Yong Jae Lee and Kristen Grauman University of Texas at Austin 1.
Context Neelima Chavali ECE /21/2013. Roadmap Introduction Paper1 – Motivation – Problem statement – Approach – Experiments & Results Paper 2 Experiments.
Object Recognition by Integrating Multiple Image Segmentations Caroline Pantofaru, Cordelia Schmid, Martial Hebert ECCV 2008 E.
Learning video saliency from human gaze using candidate selection CVPR2013 Poster.
A Framework to Predict the Quality of Answers with Non-Textual Features Jiwoon Jeon, W. Bruce Croft(University of Massachusetts-Amherst) Joon Ho Lee (Soongsil.
Real-Time Feature Matching using Adaptive and Spatially Distributed Classification Trees Aurélien BOFFY, Yanghai TSIN, Yakup GENC.
Finding Clusters within a Class to Improve Classification Accuracy Literature Survey Yong Jae Lee 3/6/08.
Learning Hierarchical Features for Scene Labeling Cle’ment Farabet, Camille Couprie, Laurent Najman, and Yann LeCun by Dong Nie.
Image segmentation.
Scene Parsing with Object Instances and Occlusion Ordering JOSEPH TIGHE, MARC NIETHAMMER, SVETLANA LAZEBNIK 2014 IEEE CONFERENCE ON COMPUTER VISION AND.
Parsing Natural Scenes and Natural Language with Recursive Neural Networks INTERNATIONAL CONFERENCE ON MACHINE LEARNING (ICML 2011) RICHARD SOCHER CLIFF.
Deeply-Recursive Convolutional Network for Image Super-Resolution
Recent developments in object detection
Guillaume-Alexandre Bilodeau
Krishna Kumar Singh, Yong Jae Lee University of California, Davis
Finding Things: Image Parsing with Regions and Per-Exemplar Detectors
Nonparametric Semantic Segmentation
Context-Aware Modeling and Recognition of Activities in Video
Object-Graphs for Context-Aware Category Discovery
Rob Fergus Computer Vision
Weakly Supervised Action Recognition
KFC: Keypoints, Features and Correspondences
Human-object interaction
“Traditional” image segmentation
Learning to Detect Human-Object Interactions with Knowledge
ICCV 2019.
Presentation transcript:

INTRODUCTION Heesoo Myeong, Ju Yong Chang, and Kyoung Mu Lee Department of EECS, ASRI, Seoul National University, Seoul, Korea Learning Object Relationships via Graph-based Context Model Why Context for Scene Understanding? PROPOSED METHOD Results on SIFT Flow Dataset Quantitative Results on Standard Datasets   Jain et al. dataset (Jain et al., ECCV10): 250 training images, 100 test images, 19 labels   SIFT Flow dataset (Liu et al., CVPR09): 2,488 training images, 200 test images, 33 labels Table 1: Per-pixel classification rates and (average per-class rates)   For a test image, retrieve T most similar training images using global features EXPERIMENTS Inference   Use Fully connected Markov Random Field (MRF) model: Previous Works & Limitations Our Contributions Prefer frequently appeared objects Not invariant to the number of pixels/regions   A novel context link view of contextual knowledge   Applying label propagation to context link prediction   Nonparametric context model scalable to large datasets   Incorporating contextual information is crucial for scene understanding   Recently, object-object relationships have shown better performance than scene- object relationships [Rabinovich et al., ICCV07] Our Approach   The similarity graph naturally reflects visual similarity Context link on the similarity graph Semi-supervised context link prediction   A novel view for representing object relationships   Object relationships are usually formulated as co-occurrence statistics or spatial relation [Rabinovich et al., ICCV07, Gould et al., IJCV08, Jain et al., ECCV10]   No appearance info. has been taken into account during context modeling process Building Building Building Building ?? Road Query image Conventional context model Building Crosswalk Our context model Building Retrieved training image (Exemplar) Crosswalk   Graph-based and Exemplar-based context model   Utilize object relationships adaptively according to the visual appearance of objects   Learn object-object relationships between all pairs of regions across whole object class pairs Retrieved training images (Exemplars) Test image (building,car)-link Similarity graph ? ? car predicted (building,car)-link Building Similarity edge Similarity edge   K-nearest neighbor similarity graph is constructed among regions from both the query image and the corresponding retrieved image set   Decompose context link prediction problem into two independent label propagation subproblems [Lu and Ip, ECCV10] Label propagation sky building car road crosswalk building car road car sidewalk building sky car plant Input imageGround truthSuperParsingOursBaseline classifier Input image Ground truthSuperParsingOursBaseline classifier tree sky bison field sky water sand Results on Jain et al. Dataset   We have proposed a novel framework for modeling image-dependent contextual relationships using graph-based context model   Experimental results demonstrate that the proposed context model overcome the limitation of conventional context models relying on object label agreement and gives richer appearance-based context information Baseline MRF Our approach (b) Jain et al. dataset (a) SIFT flow dataset Conclusion Per-class recognition rate