Coarse-to-Fine Cross-modality Generation for Enhancing Vehicle Re-Identification with High-Fidelity Synthetic Data

Authors: Leyang Jin, Wei Ji, Tat-seng Chua, Zhedong Zheng

Published in IEEE International Conference on Robotics and Automation (ICRA), 2025

Recommended citation: Leyang Jin, Wei Ji, Tat-seng Chua, Zhedong Zheng, "Coarse-to-Fine Cross-modality Generation for Enhancing Vehicle Re-Identification with High-Fidelity Synthetic Data." IEEE International Conference on Robotics and Automation, 2025.
Download PDF: https://zdzheng.xyz/files/2025/ICRA25-Vehicle.pdf

Abstract: Due to the critical issues of privacy and partial occlusion, license plate information is not always available in vehicle recognition systems. Consequently, researchers have increasingly turned towards vehicle re-identification (reID) techniques to bridge the gap between cross-view camera systems. Despite the growing interest, one major challenge persists: the scarcity of authentic, large-scale training datasets. To address this challenge, this paper introduces a coarse-to-fine generation pipeline designed to synthesize high-fidelity vehicle data, thereby facilitating subsequent vehicle representation learning. Specifically, the proposed approach consists of three stages: Prompt Processing, Diffusion Fine-tuning, and Semantic Filtering. First, we collect detailed prompts from vehicle websites and companies with fine-grained vehicle prototype attributes. Next, we leverage the prior knowledge of these automotive prototypes to fine-tune diffusion models. Finally, to ensure the quality of the synthesized data, we employ pretrained vision-language models to filter out substandard images. Building upon the high-quality data generated by this pipeline, we validate the effectiveness using vanilla models. Extensive experimental evaluations demonstrate that our approach achieves competitive accuracy on public benchmarks such as VeRi-776, VehicleID and CityFlowV2, and is compatible with various model architectures.

@inproceedings{jin2025coarse,
author = "Jin, Leyang and Ji, Wei and Chua, Tat-seng and Zheng, Zhedong",
title = "Coarse-to-Fine Cross-modality Generation for Enhancing Vehicle Re-Identification with High-Fidelity Synthetic Data",
abstract = "Due to the critical issues of privacy and partial occlusion, license plate information is not always available in vehicle recognition systems. Consequently, researchers have increasingly turned towards vehicle re-identification (reID) techniques to bridge the gap between cross-view camera systems. Despite the growing interest, one major challenge persists: the scarcity of authentic, large-scale training datasets. To address this challenge, this paper introduces a coarse-to-fine generation pipeline designed to synthesize high-fidelity vehicle data, thereby facilitating subsequent vehicle representation learning. Specifically, the proposed approach consists of three stages: Prompt Processing, Diffusion Fine-tuning, and Semantic Filtering. First, we collect detailed prompts from vehicle websites and companies with fine-grained vehicle prototype attributes. Next, we leverage the prior knowledge of these automotive prototypes to fine-tune diffusion models. Finally, to ensure the quality of the synthesized data, we employ pretrained vision-language models to filter out substandard images. Building upon the high-quality data generated by this pipeline, we validate the effectiveness using vanilla models. Extensive experimental evaluations demonstrate that our approach achieves competitive accuracy on public benchmarks such as VeRi-776, VehicleID and CityFlowV2, and is compatible with various model architectures.",
booktitle = "IEEE International Conference on Robotics and Automation",
url = "https://zdzheng.xyz/files/2025/ICRA25-Vehicle.pdf",
funding = "SRG2024-00002-FST, MYRG-GRG2024-00077-FST-UMDF",
year = "2025" }