CCUP: A Controllable Synthetic Data Generation Pipeline for Pretraining Cloth-Changing Person Re-Identification Models

2024-10-17

arXiv.org

AI Summary

To address overfitting in cloth-changing person re-identification (CC-ReID) caused by scarce real annotations and high synthetic data costs, this paper proposes a low-cost, controllable synthetic data generation paradigm coupled with a scalable pretraining-finetuning framework. We introduce CCUP—the first large-scale self-annotated CC-ReID synthetic dataset—containing 6,000 identities and 1.18 million images, enabling fine-grained clothing variation and multi-view-consistent rendering. Leveraging virtual human modeling and multi-camera synthesis, our approach achieves outfit-level control and cross-camera appearance consistency. We adopt a two-stage training strategy using TransReID and FIRe² for pretraining and finetuning, respectively. Our method achieves state-of-the-art performance on PRCC, VC-Clothes, and NKUP, improving mAP by 8.2% and 6.7% over prior works, demonstrating strong generalization and practical applicability.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Cognitive Modeling & Cognitive Systems: Affective Computing
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Cloth-changing person re-identification (CC-ReID), also known as Long-Term Person Re-Identification (LT-ReID) is a critical and challenging research topic in computer vision that has recently garnered significant attention. However, due to the high cost of constructing CC-ReID data, the existing data-driven models are hard to train efficiently on limited data, causing overfitting issue. To address this challenge, we propose a low-cost and efficient pipeline for generating controllable and high-quality synthetic data simulating the surveillance of real scenarios specific to the CC-ReID task. Particularly, we construct a new self-annotated CC-ReID dataset named Cloth-Changing Unreal Person (CCUP), containing 6,000 IDs, 1,179,976 images, 100 cameras, and 26.5 outfits per individual. Based on this large-scale dataset, we introduce an effective and scalable pretrain-finetune framework for enhancing the generalization capabilities of the traditional CC-ReID models. The extensive experiments demonstrate that two typical models namely TransReID and FIRe^2, when integrated into our framework, outperform other state-of-the-art models after pretraining on CCUP and finetuning on the benchmarks such as PRCC, VC-Clothes and NKUP. The CCUP is available at: https://github.com/yjzhao1019/CCUP.

Problem

• Generates synthetic data for cloth-changing person re-identification
• Addresses data scarcity and overfitting in CC-ReID models
• Proposes pretrain-finetune framework to enhance model generalization

Innovation

• Generates controllable synthetic CC-ReID data
• Introduces scalable pretrain-finetune framework
• Enhances generalization with large-scale dataset