LTDA-Drive: LLMs-guided Generative Models based Long-tail Data Augmentation for Autonomous Driving

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yurt, Mahmut, Ye, Xin, Ma, Yunsheng, Luo, Jingru, Mallik, Abhirup, Pauly, John, Yaman, Burhaneddin, Ren, Liu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912391857438720
author Yurt, Mahmut
Ye, Xin
Ma, Yunsheng
Luo, Jingru
Mallik, Abhirup
Pauly, John
Yaman, Burhaneddin
Ren, Liu
author_facet Yurt, Mahmut
Ye, Xin
Ma, Yunsheng
Luo, Jingru
Mallik, Abhirup
Pauly, John
Yaman, Burhaneddin
Ren, Liu
contents 3D perception plays an essential role for improving the safety and performance of autonomous driving. Yet, existing models trained on real-world datasets, which naturally exhibit long-tail distributions, tend to underperform on rare and safety-critical, vulnerable classes, such as pedestrians and cyclists. Existing studies on reweighting and resampling techniques struggle with the scarcity and limited diversity within tail classes. To address these limitations, we introduce LTDA-Drive, a novel LLM-guided data augmentation framework designed to synthesize diverse, high-quality long-tail samples. LTDA-Drive replaces head-class objects in driving scenes with tail-class objects through a three-stage process: (1) text-guided diffusion models remove head-class objects, (2) generative models insert instances of the tail classes, and (3) an LLM agent filters out low-quality synthesized images. Experiments conducted on the KITTI dataset show that LTDA-Drive significantly improves tail-class detection, achieving 34.75\% improvement for rare classes over counterpart methods. These results further highlight the effectiveness of LTDA-Drive in tackling long-tail challenges by generating high-quality and diverse data.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18198
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LTDA-Drive: LLMs-guided Generative Models based Long-tail Data Augmentation for Autonomous Driving
Yurt, Mahmut
Ye, Xin
Ma, Yunsheng
Luo, Jingru
Mallik, Abhirup
Pauly, John
Yaman, Burhaneddin
Ren, Liu
Robotics
3D perception plays an essential role for improving the safety and performance of autonomous driving. Yet, existing models trained on real-world datasets, which naturally exhibit long-tail distributions, tend to underperform on rare and safety-critical, vulnerable classes, such as pedestrians and cyclists. Existing studies on reweighting and resampling techniques struggle with the scarcity and limited diversity within tail classes. To address these limitations, we introduce LTDA-Drive, a novel LLM-guided data augmentation framework designed to synthesize diverse, high-quality long-tail samples. LTDA-Drive replaces head-class objects in driving scenes with tail-class objects through a three-stage process: (1) text-guided diffusion models remove head-class objects, (2) generative models insert instances of the tail classes, and (3) an LLM agent filters out low-quality synthesized images. Experiments conducted on the KITTI dataset show that LTDA-Drive significantly improves tail-class detection, achieving 34.75\% improvement for rare classes over counterpart methods. These results further highlight the effectiveness of LTDA-Drive in tackling long-tail challenges by generating high-quality and diverse data.
title LTDA-Drive: LLMs-guided Generative Models based Long-tail Data Augmentation for Autonomous Driving
topic Robotics
url https://arxiv.org/abs/2505.18198