BarbieGait: An Identity-Consistent Synthetic Human Dataset with Versatile Cloth-Changing for Gait Recognition

arXiv cs.CV / 4/15/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces BarbieGait, a synthetic human gait dataset that maps real individuals into a virtual engine to generate extensive clothing variations while maintaining identity-consistent gait information for cross-clothing evaluation.
By controllably generating large amounts of synthetic training/benchmark data, BarbieGait aims to address a key real-world limitation: verifying and measuring clothing-induced variations is difficult with real captures alone.
The authors identify cross-clothing gait recognition as a cloth-invariance challenge and propose GaitCLIF, a baseline model designed to learn robust cloth-invariant features.
Experiments show that GaitCLIF substantially improves cross-clothing performance on BarbieGait and also on existing popular gait recognition benchmarks.
The work positions BarbieGait and the associated modeling baseline as enabling further progress in gait recognition under clothing changes, potentially advancing research in cloth-robust biometrics.

Abstract

Gait recognition, as a reliable biometric technology, has seen rapid development in recent years while it faces significant challenges caused by diverse clothing styles in the real world. This paper introduces BarbieGait, a synthetic gait dataset where real-world subjects are uniquely mapped into a virtual engine to simulate extensive clothing changes while preserving their gait identity information. As a pioneering work, BarbieGait provides a controllable gait data generation method, enabling the production of large datasets to validate cross-clothing issues that are difficult to verify with real-world data. However, the diversity of clothing increases intra-class variance and makes one of the biggest challenges to learning cloth-invariant features under varying clothing conditions. Therefore, we propose GaitCLIF (Gait-oriented CLoth-Invariant Feature) as a robust baseline model for cross-clothing gait recognition. Through extensive experiments, we validate that our method significantly improves cross-clothing performance on BarbieGait and the existing popular gait benchmarks. We believe that BarbieGait, with its extensive cross-clothing gait data, will further advance the capabilities of gait recognition in cross-clothing scenarios and promote progress in related research.