Computer Vision - ACCV 2012 [electronic resource] : 11th Asian Conference on Computer Vision, Daejeon, Korea, November 5-9, 2012, Revised Selected Papers, Part III / edited by Kyoung Mu Lee, Yasuyuki Matsushita, James M. Rehg, Zhanyi Hu.

Oral Session 6: Optical Flow and Tracking -- Adaptive Integration of Feature Matches into Variational Optical Flow Methods -- Efficient Learning of Linear Predictors Using Dimensionality Reduction -- Robust Visual Tracking Using Dynamic Classifier Selection with Sparse Representation of Label Noise -- Poster Session 6: Motion, Tracking, and Computational Photography Dynamic Objectness for Adaptive Tracking -- Visual Tracking in Continuous Appearance Space via Sparse Coding -- Robust Object Tracking in Crowd Dynamic Scenes Using Explicit Stereo Depth -- Structured Visual Tracking with Dynamic Graph -- Online Multi-target Tracking by Large Margin Structured Learning -- An Anchor Patch Based Optimization Framework for Reducing Optical Flow Drift in Long Image Sequences -- One-Class Multiple Instance Learning and Applications to Target Tracking -- Dense Scene Flow Based on Depth and Multi-channel Bilateral Filter -- Object Tracking within the Framework of Concept Drift -- Multiple Target Tracking Using Frame Triplets -- Spatio-Temporal Clustering Model for Multi-object Tracking through Occlusions -- Robust Object Tracking Using Constellation Model with Superpixel -- Robust Registration-Based Tracking by Sparse Representation with Model Update -- Robust and Efficient Pose Estimation from Line Correspondences -- Nonlocal Spectral Prior Model for Low-Level Vision -- Simultaneous Multiple Rotation Averaging Using Lagrangian Duality -- Observation-Driven Adaptive Differential Evolution for Robust Bronchoscope 3-D Motion Tracking -- Tracking Growing Axons by Particle Filtering in 3D + t Fluorescent Two-Photon Microscopy Images -- Image Upscaling Using Multiple Dictionaries of Natural Image Patches -- A Biologically Motivated Double-Opponency Approach to Illumination Invariance -- Measuring Linearity of Closed Curves and Connected Compound Curves -- Patch Mosaic for Fast Motion Deblurring -- Single-Image Blind Deblurring for Non-uniform Camera-Shake Blur -- Image Super-Resolution Using Local Learnable Kernel Regression -- MRF-Based Blind Image Deconvolution -- Efficient Image Appearance Description Using Dense Sampling Based Local Binary Patterns -- Navigation toward Non-static Target Object Using Footprint Detection Based Tracking -- Single Image Super Resolution Reconstruction in Perturbed Exemplar Sub-space -- Image Super-Resolution: Use of Self-learning and Gabor Prior -- Oral Session 7: Video Analysis and Action -- Recognition Action Disambiguation Analysis Using Normalized Google-Like Distance Correlogram -- Alpha-Flow for Video Matting -- Combinational Subsequence Matching for Human Identification from General Actions -- Poster Session 7: Video Analysis and Action Recognition Iterative Semi-Global Matching for Robust Driver Assistance Systems -- Action Recognition Using Canonical Correlation Kernels -- A New Framework for Background Subtraction Using Multiple Cues -- Weighted Interaction Force Estimation for Abnormality Detection in Crowd Scenes -- Egocentric Activity Monitoring and Recovery -- Spatiotemporal Salience via Centre-Surround Comparison of Visual Spacetime Orientations -- Temporal-Spatial Refinements for Video Concept Fusion -- Features with Feelings-Incorporating User Preferences in Video Categorization -- A Comparative Study of Encoding, Pooling and Normalization Methods for Action Recognition -- Dynamic Saliency Models and Human Attention: A Comparative Study on Videos -- Horror Video Scene Recognition Based on Multi-view Multi-instance Learning -- Learning Object Appearance from Occlusions Using Structure and Motion Recovery -- Exploring the Similarities of Neighboring Spatiotemporal Points for Action Pair Matching -- Sequential Reconstruction Segment-Wise Feature Track and Structure Updating Based on Parallax Paths -- Generic Active Appearance Models Revisited -- Tracking Pedestrian with Multi-component Online Deformable Part-Based Model -- Local Distance Comparison for Multiple-shot People Re-identification -- Non-sequential Multi-view Detection, Localization and Identification of People Using Multi-modal Feature Maps -- Full 6DOF Pose Estimation from Geo-Located Images -- Learning a Quality-Based Ranking for Feature Point Trajectories.