超越物体：面向细粒度分类的上下文合成数据生成
Beyond Objects: Contextual Synthetic Data Generation for Fine-Grained Classification
BOB 提出了一种微调 T2I 模型以生成高质量合成训练数据的策略：通过类无关属性（背景、姿态）显式条件化并在生成时边缘化，缓解过拟合并保留生成先验，可直接用于低样本细粒度分类的数据增强。
原文依据 We propose a fine-tuning strategy BOB (Beyond OBjects) for mitigating these concerns.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论文本到图像（T2I）模型越来越多地用于合成数据集生成，但生成合成训练数据以提升细粒度分类性能仍具挑战性。用少量真实样本微调 T2I 模型有助于生成更合适的合成训练数据；然而，这种微调也可能引入过拟合并降低生成样本的多样性。我们提出一种微调策略 BOB（Beyond OBjects）以缓解这些问题。给定一小组真实样本，我们首先使用类无关属性（如场景背景和物体姿态）对其进行描述。然后，在微调 T2I 模型时显式地以这些属性为条件，并在生成时将它们边缘化。该设计缓解过拟合，从而保留 T2I 模型的生成先验并减少估计误差，并进一步最小化意外的类间关联。跨多个 T2I 模型、骨干网络和数据集的广泛实验表明，在合成数据增强下，低样本细粒度分类达到最优性能。具体而言，BOB 在 Aircraft 数据集上比 DataDream 高出 7.4%（当用 5 张真实图像增强 100 张合成图像微调 CLIP 分类器时，从 50.0% 提升至 57.4%）。此外，在四个数据集中的三个上，使用 BOB 生成的合成数据和五张真实图像微调下游模型，其性能优于使用十张真实图像微调。总体而言，BOB 在 24 个实验设置中的 18 个上优于先前技术，其中 14 个设置准确率提升超过 2%。
英文原文摘要 1645 字符
Text-to-image (T2I) models are increasingly used for synthetic dataset generation, but generating synthetic training data to improve fine-grained classification performance remains challenging. Fine-tuning the T2I model with a few real examples can help generate more appropriate synthetic training data; however, this fine-tuning may also introduce overfitting and reduce diversity in the generated samples. We propose a fine-tuning strategy BOB (Beyond OBjects) for mitigating these concerns. Given a small set of real examples, we first describe them using class-agnostic attributes such as scene background and object pose. We then explicitly condition on these attributes during fine-tuning of the T2I model and marginalize them out during generation. This design mitigates overfitting, thus preserving the T2I model's generative prior and reducing estimation errors, and further minimizes unintended inter-class associations. Extensive experiments across multiple T2I models, backbones, and datasets demonstrate state-of-the-art performance in low-shot fine-grained classification when augmented with synthetic data. Concretely, BOB outperforms DataDream by 7.4% on the Aircraft dataset (from 50.0% to 57.4% when fine-tuning a CLIP classifier with 5 real images augmented with 100 synthetic images). Additionally, in three of the four datasets, the fine-tuning downstream models with synthetic data generated from BOB and five real images achieves better performance than fine-tuning with 10 real images. Collectively, BOB outperforms prior art in 18 of 24 experimental settings, with over 2% accuracy improvements in 14 of these settings.