附註:Includes bibliographical references and index.
Learning and Recognition -- A Constrained Semi-supervised Learning Approach to Data Association -- Learning Mixtures of Weighted Tree-Unions by Minimizing Description Length -- Decision Theoretic Modeling of Human Facial Displays -- Kernel Feature Selection with Side Data Using a Spectral Approach -- Tracking II -- Tracking Articulated Motion Using a Mixture of Autoregressive Models -- Novel Skeletal Representation for Articulated Creatures -- An Accuracy Certified Augmented Reality System for Therapy Guidance -- Posters III -- 3D Human Body Tracking Using Deterministic Temporal Motion Models -- Robust Fitting by Adaptive-Scale Residual Consensus -- Causal Camera Motion Estimation by Condensation and Robust Statistics Distance Measures -- An Adaptive Window Approach for Image Smoothing and Structures Preserving -- Extraction of Semantic Dynamic Content from Videos with Probabilistic Motion Models -- Are Iterations and Curvature Useful for Tensor Voting? -- A Feature-Based Approach for Determining Dense Long Range Correspondences -- Combining Geometric- and View-Based Approaches for Articulated Pose Estimation -- Shape Matching and Recognition -- Using Generative Models and Informative Features -- Generalized Histogram: Empirical Optimization of Low Dimensional Features for Image Matching -- Recognizing Objects in Range Data Using Regional Point Descriptors -- Shape Reconstruction from 3D and 2D Data Using PDE-Based Deformable Surfaces -- Structure and Motion Problems for Multiple Rigidly Moving Cameras -- Detection and Tracking Scheme for Line Scratch Removal in an Image Sequence -- Color Constancy Using Local Color Shifts -- Image Anisotropic Diffusion Based on Gradient Vector Flow Fields -- Optimal Importance Sampling for Tracking in Image Sequences: Application to Point Tracking -- Learning to Segment -- MCMC-Based Multiview Reconstruction of Piecewise Smooth Subdivision Curves with a Variable Number of Control Points -- Bayesian Correction of Image Intensity
摘要:The four-volume set comprising LNCS volumes 3021/3022/3023/3024 constitutes the refereed proceedings of the 8th European Conference on Computer Vision, ECCV 2004, held in Prague, Czech Republic, in May 2004. The 190 revised papers presented were carefully reviewed and selected from a total of 555 papers submitted. The four books span the entire range of current issues in computer vision. The papers are organized in topical sections on tracking; feature-based object detection and recognition; geometry; texture; learning and recognition; information-based image processing; scale space, flow, and restoration; 2D shape detection and recognition; and 3D shape representation and reconstruction.
