Key-Frame-Aware Hierarchical Learning for Robust Gait Recognition

Gait recognition in unconstrained environments is severely hampered by variations in view, clothing, and carrying conditions. To address this, we introduce HierarchGait, a key-frame-aware hierarchical learning framework. Our approach uniquely integrates three complementary modules: a TemplateBlock-based Motion Extraction (TBME) for coarse-to-fine anatomical feature learning, a Sequence-Level Spatio-temporal Feature Aggregator (SSFA) to identify and prioritize discriminative key-frames, and a Frame-level Feature Re-segmentation Extractor (FFRE) to capture fine-grained motion details. This synergistic design yields a robust and comprehensive gait representation. We demonstrate the superiority of our method through extensive experiments. On the highly challenging CASIA-B dataset, HierarchGait achieves new state-of-the-art average Rank-1 accuracies of 98.1% under Normal (NM), 95.9% under Bag (BG), and 87.5% under Coat (CL) conditions. Furthermore, on the large-scale OU-MVLP dataset, our model attains a 91.5% average accuracy. These results validate the significant advantage of explicitly modeling anatomical hierarchies and temporal key-moments for robust gait recognition.

• Publication year

  2025

• Venue

  Journal of Imaging

• Publication date

  2025-11-01

• Fields of study

  Medicine, Computer Science, Engineering

• Identifiers
  DOI 10.3390/jimaging11110402PMID 41295119PMCID 12653630
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Accuracy of unilateral and bilateral gait assessment using a mobile gait analysis system at different walking speeds.

  2024cited by this paper
• GaitFormer: Leveraging dual-stream spatial-temporal Vision Transformer via a single low-cost RGB camera for clinical gait analysis

  2024cited by this paper
• A Comprehensive Study on the Evaluation of Silhouette-Based Gait Recognition

  2023cited by this paper
• MetaGait: Learning to Learn an Omni Sample Adaptive Representation for Gait Recognition

  2023cited by this paper
• STAR: Spatio-Temporal Augmented Relation Network for Gait Recognition

  2023influential reference
• Gait recognition using free-area transformer networks

  2023cited by this paper
• GaitASMS: gait recognition by adaptive structured spatial representation and multi-scale temporal aggregation

  2023influential reference
• GaitEdge: Beyond Plain End-to-end Gait Recognition for Better Practicality

  2022cited by this paper
• TransGait: Multimodal-based gait recognition with set transformer

  2022cited by this paper
• Gait recognition based on vision systems: A systematic survey

  2021cited by this paper
• 3D Local Convolutional Neural Networks for Gait Recognition

  2021influential reference
• Context-Sensitive Temporal Feature Learning for Gait Recognition

  2021cited by this paper
• End-to-end Model-based Gait Recognition using Synchronized Multi-view Pose Constraint

  2021cited by this paper
• Set Residual Network for Silhouette-Based Gait Recognition

  2021cited by this paper
• Graph-optimized coupled discriminant projections for cross-view gait recognition

  2021cited by this paper
• Deep Gait Recognition: A Survey

  2021cited by this paper
• View-Invariant Gait Recognition With Attentive Recurrent Learning of Partial Representations

  2020cited by this paper
• Cross-View Gait Recognition by Discriminative Feature Learning

  2020cited by this paper
• Leader-Based Multi-Scale Attention Deep Architecture for Person Re-Identification

  2020cited by this paper
• A model-based Gait Recognition Method based on Gait Graph Convolutional Networks and Joints Relationship Pyramid Mapping

  2020cited by this paper
• GaitPart: Temporal Part-Based Model for Gait Recognition

  2020influential reference
• Performance Evaluation of Model-Based Gait on Multi-View Very Large Population Database With Pose Sequences

  2020cited by this paper
• Gait Lateral Network: Learning Discriminative and Compact Representations for Gait Recognition

  2020cited by this paper
• Gait Recognition with Multiple-Temporal-Scale 3D Convolutional Neural Network

  2020cited by this paper
• Gait Recognition via Effective Global-Local Feature Representation and Local Temporal Aggregation

  2020influential reference
• Learning Joint Gait Representation via Quintuplet Loss Minimization

  2019cited by this paper
• Human Gait Recognition Based on Frame-by-Frame Gait Energy Images and Convolutional Long Short-Term Memory

  2019cited by this paper
• Gait Recognition via Disentangled Representation Learning

  2019cited by this paper
• Gait and jump classification in modern equestrian sports

  2018cited by this paper
• GaitSet: Regarding Gait as a Set for Cross-View Gait Recognition

  2018cited by this paper
• Multi-view large population gait dataset and its performance evaluation for cross-view gait recognition

  2018cited by this paper
• Robust gait recognition: a comprehensive survey

  2018cited by this paper
• Person Re-identification by Deep Learning Multi-scale Representations

  2017cited by this paper
• In Defense of the Triplet Loss for Person Re-Identification

  2017influential reference
• Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification

  2017cited by this paper
• Beyond Part Models: Person Retrieval with Refined Part Pooling

  2017cited by this paper
• View-invariant Gait Recognition Using Convolutional Neural Network ∗

  2016cited by this paper
• Clothing-invariant gait recognition using convolutional neural network

  2016cited by this paper
• Design of intelligent recognition system based on gait recognition technology in smart transportation

  2016cited by this paper
• Multi-view gait recognition using 3D convolutional neural networks

  2016cited by this paper
• Gait Recognition: Databases, Representations, and Applications

  2015cited by this paper
• Gait Verification System for Supporting Criminal Investigation

  2013cited by this paper
• Gait recognition using Gait Entropy Image

  2009cited by this paper
• View Invariant Gait Recognition

  2009cited by this paper
• A Framework for Evaluating the Effect of View Angle, Clothing and Carrying Condition on Gait Recognition

  2006cited by this paper
• Individual recognition using gait energy image

  2006cited by this paper

• No citing papers are available for this paper.