LP$^{2}$DH: A Locality-Preserving Pixel-Difference Hashing Framework for Dynamic Texture Recognition

2026-04-17

AI Summary

This work addresses the limitations of Spatiotemporal Local Binary Patterns (ST-LBP) in dynamic texture recognition—namely, their high feature dimensionality and neglect of correlations across spatiotemporal planes—by proposing a local-structure-preserving pixel-difference hashing framework. The method jointly models pixel difference vectors over the full spatiotemporal neighborhood for the first time, and co-optimizes the hashing matrix and compact binary codes via curvature-aware gradient descent on the Stiefel manifold. A discriminative feature representation is further enhanced through dictionary learning. Evaluated on three standard benchmarks—UCLA, DynTex++, and YUPENN—the approach achieves state-of-the-art accuracy rates of 99.80%, 98.52%, and 96.19%, respectively, significantly outperforming existing methods.

Technology Category

• Machine Learning: Learning with Manifolds
• Computer Vision: Representation Learning for Vision
• Search and Optimization: Local Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Spatiotemporal Local Binary Pattern (STLBP) is a widely used dynamic texture descriptor, but it suffers from extremely high dimensionality. To tackle this, STLBP features are often extracted on three orthogonal planes, which sacrifice inter-plane correlation. In this work, we propose a Locality-Preserving Pixel-Difference Hashing (LP$^{2}$DH) framework that jointly encodes pixel differences in the full spatiotemporal neighbourhood. LP$^{2}$DH transforms Pixel-Difference Vectors (PDVs) into compact binary codes with maximal discriminative power. Furthermore, we incorporate a locality-preserving embedding to maintain the PDVs' local structure before and after hashing. Then, a curvilinear search strategy is utilized to jointly optimize the hashing matrix and binary codes via gradient descent on the Stiefel manifold. After hashing, dictionary learning is applied to encode the binary vectors into codewords, and the resulting histogram is utilized as the final feature representation. The proposed LP$^{2}$DH achieves state-of-the-art performance on three major dynamic texture recognition benchmarks: 99.80% against DT-GoogleNet's 98.93% on UCLA, 98.52% against HoGF$^{3D}$'s 97.63% on DynTex++, and 96.19% compared to STS's 95.00% on YUPENN. The source code is available at: https://github.com/drx770/LP2DH.

Problem

• dynamic texture recognition
• STLBP
• high dimensionality
• inter-plane correlation
• feature representation

Innovation

• Locality-Preserving Hashing
• Pixel-Difference Vectors
• Dynamic Texture Recognition
• Stiefel Manifold Optimization
• Binary Code Learning