ICPR 2010

383 papers
The 2010 edition of the International Conference on Pattern Recognition (ICPR) featured 488 papers covering a broad spectrum of computer vision and pattern recognition topics. Research themes focused on enhancing the robustness of fundamental tasks, such as boundary optimization, interactive segmentation, and visual tracking. Key efforts were directed toward mitigating temporal drift in video propagation and addressing the shrinking bias in boundary optimization. The year also saw a shift toward scaling semantic segmentation from binary to multi-class environments and achieving sub-pixel precision through continuous geometric modeling. Human-centric analysis was a prominent theme, with studies exploring group activity recognition using Gaussian processes, action recognition through spatial-temporal context, and affective behavior analysis in uncontrolled environments. Other research areas included the detection of natural scene text in video, face verification, and the disambiguation of repetitive structures for 3D reconstruction. Prolific contributors to this year's proceedings included Horst Bischof, Edwin R. Hancock, and Qingming Huang. These technical contributions emphasized stability and consistency, particularly in tracking and segmentation, reflecting a focus on improving performance in complex visual settings.

Top Research Directions

Research directions most represented by ICPR 2010 papers.

Frequency Domain Mapping for Geometric Invariance 2
Generalizing Emotion Recognition to Unconstrained Environments 2
Constraining Geometric Search with Local Gradient Directions 1
Statistical Stacking for Signal Enhancement 1
Continuous Domain Optimization for Subpixel Boundary Precision 1
Overcoming Metrication Errors in Digital Geometry 1
Sub-Pixel Refinement for Geometric Precision 1
Finite Dimensional Reductions via the Representer Theorem 1
Spatial Prompt Sensitivity in Interactive Segmentation 1
Stabilizing Medial Axis Transforms Against Boundary Noise 1
Eliminating Manual Landmark Dependency in Shape Alignment 1
Motion-Appearance Fusion in Video Object Tracking and Segmentation 1
Robust Visual Matching via Gradient Orientation Invariance 1
Localized Feature Extraction for Robust Facial Analysis 1
Biometric Recognition under Severe Resolution Degradation 1

Top Authors

AuthorPapers
Horst Bischof9
Edwin Hancock8
Yunde Jia6
Brian C. Lovell5
Qingming Huang5
Vittorio Murino5
Cuntai Guan4
Kyoung Mu Lee4
Shuqiang Jiang4
Tieniu Tan4
Venu Govindaraju4
Weiming Hu4
Xiaochun Cao4
Xinge You4
Alberto Del Bimbo3
Christopher Leckie3
Ioannis Patras3
Joni-kristian Kamarainen3
Krystian Mikolajczyk3
Liming Chen3

Impactful Papers

Papers from ICPR 2010 that shaped research clusters.

Foundational

Influential

Papers

Title
Suggesting Songs for Media Creation Using Semantics
Symbol Classification Using Dynamic Aligned Shape Descriptor
Task-Oriented Evaluation of Super-Resolution Techniques
Text Separation from Mixed Documents Using a Tree-Structured Classifier
The Balanced Accuracy and Its Posterior Distribution
The Binormal Assumption on Precision-Recall Curves
The Fusion of Deep Learning Architectures and Particle Filtering Applied to Lip Tracking
The Human Action Image
The Impact of Color on Bag-of-Words Based Object Recognition
Three-layer Spatial Sparse Coding for Image Classification
Time Series Classification Using Support Vector Machine with Gaussian Elastic Metric Kernel
Topic-Sensitive Tag Ranking
Torchlight Navigation
Tracking Multiple People with Illumination Maps
Tracking Objects with Adaptive Feature Patches for PTZ Camera Visual Surveillance
Triangle-Constraint for Finding More Good Features
Unified Approach to Detection and Identification of Commercial Films by Temporal Occurrence Pattern
Unifying Approach for Fast License Plate Localization and Super-Resolution
Unsupervised Ensemble Ranking: Application to Large-Scale Image Retrieval
Unsupervised Learning from Linked Documents
Unsupervised Learning of Stroke Tagger for Online Kanji Handwriting Recognition
Unsupervised Visual Object Categorisation via Self-organisation
User Adaptive Clustering for Large Image Databases
Using Spatial Relations for Graphical Symbol Description
Using Symmetry to Select Fixation Points for Segmentation
Video Retrieval Based on Tracked Features Quantization
View Invariant Body Pose Estimation Based on Biased Manifold Learning
Visibility of Multiple Cameras in a Scene with Unknown Geometry
Visual Gaze Estimation by Joint Head and Eye Information
Visual Tracking Using Sparsity Induced Similarity
Water Reflection Detection Using a Flip Invariant Shape Detector
Wavelet Domain Local Binary Pattern Features For Writer Identification
Weakly Supervised Action Recognition Using Implicit Shape Models

« prev Page 8/8