LTDA-Drive: LLM誘導生成モデルによる自動運転向けロングテールデータ拡張
LTDA-Drive: LLMs-guided Generative Models based Long-tail Data Augmentation for Autonomous Driving
LLMが誘導する生成モデルで自動運転シーンのロングテール物体を合成し、希少クラスの検出精度を大幅に改善するデータ拡張手法を提案。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Mahmut Yurt, Xin Ye, Yunsheng Ma, Jingru Luo, Abhirup Mallik, John Pauly, Burhaneddin Yaman, Liu Ren
分類: cs.RO
原文アブストラクト
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.
