Computer Vision -- ECCV 2014

The seven-volume set comprising LNCS volumes 8689-8695 constitutes the refereed proceedings of the 13th European Conference on Computer Vision, ECCV 2014, held in Zurich, Switzerland, in September 2014.
The 363 revised papers presented were carefully reviewed and selected from 1444 submissions. The papers are organized in topical sections on tracking and activity recognition; recognition; learning and inference; structure from motion and feature matching; computational photography and low-level vision; vision; segmentation and saliency; context and 3D scenes; motion and 3D scene analysis; and poster sessions.

Coarse-to-Fine Auto-Encoder Networks (CFAN) for Real-Time Face Alignment.- From Manifold to Manifold: Geometry-Aware Dimensionality Reduction for SPD Matrices.- Pose Machines: Articulated Pose Estimation via Inference Machines.- Piecewise-Planar StereoScan: Structure and Motion from Plane Primitives.- Nonrigid Surface Registration and Completion from RGBD Images.- Unsupervised Dense Object Discovery, Detection, Tracking and Reconstruction.- Know Your Limits: Accuracy of Long Range Stereoscopic Object Measurements in Practice.- As-Rigid-As-Possible Stereo under Second Order Smoothness Priors.- Real-Time Minimization of the Piecewise Smooth Mumford-Shah Functional.- A MAP-Estimation Framework for Blind Deblurring Using High-Level Edge Priors.- Efficient Color Constancy with Local Surface Reflectance Statistics.- A Contrast Enhancement Framework with JPEG Artifacts Suppression.- Radial Bright Channel Prior for Single Image Vignetting Correction.- Tubular Structure Filtering by Ranking Orientation Responses of Path Operators.- Optimization-Based Artifact Correction for Electron Microscopy Image Stacks.- Metric-Based Pairwise and Multiple Image Registration.- Canonical Correlation Analysis on Riemannian Manifolds and Its Applications.- Scalable 6-DOF Localization on Mobile Devices.- On Mean Pose and Variability of 3D Deformable Models.- Hybrid Stochastic / Deterministic Optimization for Tracking Sports Players and Pedestrians.- What Do I See? Modeling Human Visual Perception for Multi-person Tracking.- Consistent Re-identification in a Camera Network.- Surface Normal Deconvolution: Photometric Stereo for Optically Thick Translucent Objects.- Intrinsic Video.- Robust and Accurate Non-parametric Estimation of Reflectance Using Basis Decomposition and Correction Functions.- Intrinsic Textures for Relightable Free-Viewpoint Video.- Reasoning about Object Affordances in a Knowledge Base Representation.- Binary Codes Embedding for Fast Image Tagging with Incomplete Labels.- Recognizing Products: A Per-exemplar Multi-label Image Classification Approach.- Part-Pair Representation for Part Localization.- Weakly Supervised Learning of Objects, Attributes and Their Associations.- Interestingness Prediction by Robust Learning to Rank.- Pairwise Probabilistic Voting: Fast Place Recognition without RANSAC.- Robust Instance Recognition in Presence of Occlusion and Clutter.- Learning 6D Object Pose Estimation Using 3D Object Coordinates.- Growing Regression Forests by Classification: Applications to Object Pose Estimation.- Stacked Deformable Part Model with Shape Regression for Object Part Localization.- Transductive Multi-view Embedding for Zero-Shot Recognition and Annotation.- Self-explanatory Sparse Representation for Image Classification.- Efficient k-Support Matrix Pursuit.- Geodesic Regression on the Grassmannian.- Model Selection by Linear Programming.- Perceptually Inspired Layout-Aware Losses for Image Segmentation.- Large Margin Local Metric Learning.- Movement Pattern Histogram for Action Recognition and Retrieval.- Pose Filter Based Hidden-CRF Models for Activity Detection.- Action Recognition Using Super Sparse Coding Vector with Spatio-temporal Awareness.- HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition.- Natural Action Recognition Using Invariant 3D Motion Encoding.- Detecting Social Actions of Fruit Flies.- Progressive Mode-Seeking on Graphs for Sparse Feature Matching.- Globally Optimal Inlier Set Maximization with Unknown Rotation and Focal Length.- Match Selection and Refinement for Highly Accurate Two-View Structure from Motion.- LSD-SLAM: Large-Scale Direct Monocular SLAM.