Abstract:Generative dynamic texture models (GDTMs) are widely used for dynamic texture (DT) segmentation in the video sequences. GDTMs represent DTs as a set of linear dynamical systems (LDSs). A major limitation of these models concerns the automatic selection of a proper number of DTs. Dirichlet process mixture (DPM) models which have appeared recently as the cornerstone of the non-parametric Bayesian statistics, is an optimistic candidate toward resolving this issue. Under this motivation to resolve the aforementioned drawback, we propose a novel non-parametric fully Bayesian approach for DT segmentation, formulated on the basis of a joint DPM and GDTM construction. This interaction causes the algorithm to overcome the problem of automatic segmentation properly. We derive the Variational Bayesian Expectation-Maximization (VBEM) inference for the proposed model. Moreover, in the E-step of inference, we apply Rauch-Tung-Striebel smoother (RTSS) algorithm on Variational Bayesian LDSs. Ultimately, experiments on different video sequences are performed. Experiment results indicate that the proposed algorithm outperforms the previous methods in efficiency and accuracy noticeably.

What problem does this paper attempt to address?

This paper attempts to solve a key problem in dynamic texture (DT) segmentation, that is, how to automatically select an appropriate number of dynamic textures. Traditional generative dynamic texture models (GDTMs) need to pre - determine the number of dynamic textures when dealing with dynamic texture segmentation in video sequences, which limits the flexibility and automation level of their applications. In addition, when different frames contain different numbers of dynamic textures, this preset becomes particularly difficult. To solve this problem, the author proposes a brand - new non - parametric fully Bayesian infinite generative model (IGDTM) based on the Dirichlet process (DP). By combining GDTM and DP, this model can automatically determine the number of dynamic textures in a video sequence without human intervention. Specifically, the main contributions of this model include: 1. **Non - parametric GDTM**: The non - parametric GDTM based on the Dirichlet process is introduced to solve the problem of automatic selection of the number of dynamic textures. 2. **Fully Bayesian generative model**: A fully Bayesian generative model is proposed, which reduces the sensitivity of GDTM to parameter initialization. To achieve this goal, the author uses the variational Bayesian expectation - maximization (VBEM) algorithm for inference and applies the Rauch - Tung - Striebel smoother (RTSS) algorithm in the E - step to calculate the expected value of the variational Bayesian linear dynamic system. Experimental results show that this method is significantly superior to existing methods in terms of efficiency and accuracy. In summary, the main purpose of this paper is to solve the problem of automatic selection of the number of dynamic textures in dynamic texture segmentation by introducing a non - parametric Bayesian infinite generative model, thereby improving the accuracy and robustness of segmentation.

A Fully Bayesian Infinite Generative Model for Dynamic Texture Segmentation

Unsupervised Dynamic Texture Segmentation Using Appearance and Motion

Unsupervised Dynamic Texture Segmentation Using Local Spatiotemporal Descriptors

Dynamic texture recognition with video set based collaborative representation

Automatic Dynamic Texture Segmentation Using Local Descriptors and Optical Flow

An Improved Local Descriptor and Threshold Learning for Unsupervised Dynamic Texture Segmentation

Unsupervised Texture Segmentation Via Applying Geodesic Active Regions To Gaborian Feature Space

Dynamic Texture Recognition using PDV Hashing and Dictionary Learning on Multi-scale Volume Local Binary Pattern

Spatio-Temporal Segmentation with Depth-Inferred Videos of Static Scenes

Dynamic Facial Expression Recognition Using A Bayesian Temporal Manifold Model

Unsupervised Texture Segmentation Based on Geodesic Active Regions

A Pixel-Level Segmentation-Synthesis Framework for Dynamic Texture Video Compression

Dynamic texture analysis with diffusion in networks

Generative Image Dynamics

Dynamic Gaussian Mixture based Deep Generative Model For Robust Forecasting on Sparse Multivariate Time Series

Effective Bayesian Inference by Data-Driven Markov Chain Monte Carlo for Object Recognition and Image Segmentation

Learning Dynamic Generator Model by Alternating Back-Propagation through Time

A Novel Active Contour Model for Texture Segmentation

Image segmentation by data-driven markov chain monte carlo

Fine Detailed Texture Learning for 3D Meshes with Generative Models

Spatiotemporal Generative Adversarial Network-Based Dynamic Texture Synthesis for Surveillance Video Coding