Dynamic texture recognition

By Payam A. Saisan, Gianfranco Doretto, Ying Nian Wu, Stefano Soatto. Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, 2001

377 citations recorded by the source corpus.

Dynamic textures are sequences of images that exhibit some form of temporal stationarity, such as waves, steam, and foliage. We pose the problem of recognizing and classifying dynamic textures in the space of dynamical systems where each dynamic texture is uniquely represented. Since the space is non-linear, a distance between models must be defined We examine three different distances in the space of autoregressive models and assess their power.

Subfields: Computer Vision and Pattern Recognition, Geometry and Topology

View Paper: https://www.semanticscholar.org/paper/1ab4ac0a0e23947fcccf93c3ca26be700f0c361f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Geometry and Topology

Citations in the corpus

Dynamic Textures (2003, background): “On recognition, we show how textures alike tend to cluster in model space, therefore assessing the potential to build a recognition system based on this framework (Saisan et al., 2001).”
Learning LBP structure by maximizing the conditional mutual information (2015, methodology): “Dynamic Texture Recognition on the UCLA Database The UCLA dynamic texture database [58, 59] has been widely used as a benchmark dataset for DT categorization [4, 50, 60, 61].”
Maximum Margin Distance Learning for Dynamic Texture Recognition (2010, methodology): “We first learn class-independent and class-dependent weights for the UCLA benchmark dataset [5].”
Unsupervised Neural Learning on Lie Group (2002, background): “…the necessity of de(cid:12)ning a measure of distance between two ARMA models of that kind, which is a fundamental requirement for classi(cid:12)cation purposes, 50 it can be proven that selecting the realization that makes matrix C orthonormal gives rise to a really well-suited representation.”
Modeling Video Dynamics with Deep Dynencoder (2014, methodology): “Similar with [6][18][10], we define a distance measure between two videos based on our model and apply it to traffic scene classification and motion segmentation.”
Graphcut textures: image and video synthesis using graph cuts (2003, methodology): “Parametric models cannot synthesize as large a variety of textures as other models described here, but provide better model generalization and are more amenable to introspection and recognition [Saisan et al. 2001].”
Locally time-invariant models of human activities using trajectories on the grassmannian (2009, methodology): “When the model parameters A,C,Q,R are constant, the model reduces to the well-known time-invariant LDS which has been successfully applied in several vision tasks [16, 18].”
Human gesture recognition on product manifolds (2012, background): “Some of the earliest works involved dynamic texture recognition (Saisan et al., 2001) and human gait recognition (Bissacco et al., 2001).”
Canonical Correlation Analysis of Video Volume Tensors for Action Categorization and Detection (2009, background): “Keywords– Action categorization, gesture recognition, canonical correlation analysis, tensor, action detection, incremental subspace learning, spatiotemporal pattern classification.”
Higher Order SVD Analysis for Dynamic Texture Synthesis (2008, methodology): “Table I summarizes the different methods for analysis/synthesis of dynamic textures, highlighting their advantages and drawbacks.”
DynTex: A comprehensive database of dynamic textures (2010, background): “Research on texture synthesis aims to model dynamic textures for realistic rendering (e.g. Szummer and Picard, 1996, Saisan et al., 2001 or Filip et al., 2006).”
Analysis and synthesis of textured motion: particles and waves (2004, background): “Index Terms—Textured motion, generative model, texton, statistical learning, object tracking, stochastic gradient.”
Event-based analysis of video (2001, background): “Analysis of events [27, 2, 10, 16, 15, 7, 18, 11, 21] has primarily focused on the recognition of sets of predefined events or actions, or assumed restricted imaging environments.”