GaitKD: A Universal Decoupled Distillation Framework for Efficient Gait Recognition

2026-04-28

AI Summary

This work addresses the high computational cost of state-of-the-art gait recognition models and the limited efficacy of conventional knowledge distillation when applied to part-structured gait architectures. To overcome these challenges, the authors propose a decoupled knowledge distillation framework that, for the first time, decomposes gait knowledge transfer into two complementary components: decision-level and boundary-level distillation. The former preserves discriminative decision relationships through part-calibrated logit distillation, while the latter maintains class-boundary structures in the embedding space via activation boundary alignment. The framework accommodates heterogeneous teacher–student architectures and incurs no additional inference overhead. Extensive experiments demonstrate that the proposed method significantly outperforms strong baselines across multiple gait recognition benchmarks, confirming that boundary-preserving distillation offers superior stability and effectiveness compared to direct feature regression.

Abstract

Gait recognition is an attractive biometric modality for long-range and contact-free identification, but high-performing gait models often rely on deep and computationally expensive architectures that are difficult to deploy in practice. Knowledge distillation (KD) offers a natural way to transfer knowledge from a powerful teacher to an efficient student; however, standard KD is often less effective for part-structured gait models, where supervision is formed from both part-wise classification logits and part-wise retrieval embeddings. In this paper, we propose GaitKD, a distillation framework that decouples gait knowledge transfer into two complementary components: decision-level distillation and boundary-level distillation. Specifically, GaitKD aligns the teacher and student through part-calibrated logit distillation to transfer inter-class decision relations, while preserving the teacher-induced partitioning of the embedding space through an activation-boundary objective instead of direct feature regression. With a simple aligned part-wise design, GaitKD supports heterogeneous teacher-student gait models without introducing additional inference cost. Experimental results across multiple gait recognition benchmarks and teacher-student configurations show consistent improvements over strong gait baselines. Our study demonstrates that the two transfer components are complementary, and boundary-preserving distillation provides more stable performance than direct feature regression. Source code is available at https://github.com/liyiersan/GaitKD/

Problem

• gait recognition
• knowledge distillation
• part-structured models
• model deployment
• biometric identification

Innovation

• gait recognition
• knowledge distillation
• decoupled distillation
• boundary-level distillation
• part-wise modeling