In this paper, considering the balance of data/model privacy of model owners and user needs, we propose a new setting called Back-Propagated Black-Box Adaptation (BPBA) for users to better train their private models via the guidance of the back-propagated results of a Black-box foundation/source model. Our setting can ease the usage of foundation/source models as well as prevent the leakage and misuse of foundation/source models. Moreover, we also propose a new training strategy called Bootstrap The Original Latent (BTOL) to fully utilize the foundation/source models. Our strategy consists of a domain adapter and a freeze-and-thaw strategy. We apply our BTOL under BPBA and Black-box UDA settings on three different datasets. Experiments show that our strategy is efficient and robust in various settings without manual augmentations.

翻译：本文中，考虑到模型所有者的数据/模型隐私与用户需求之间的平衡，我们提出了一种称为反向传播黑盒适应（BPBA）的新设置，使用户能够通过黑盒基础/源模型的反向传播结果指导其私有模型的训练。该设置既简化了基础/源模型的使用，又防止了基础/源模型的泄露和滥用。此外，我们还提出了一种名为"引导原始潜在特征"（BTOL）的新训练策略，以充分利用基础/源模型。该策略包含一个域适配器和一个冻结-解冻机制。我们将BTOL应用于BPBA和黑盒无监督域适应（UDA）设置，并在三个不同数据集上进行了实验。结果表明，在无需手动增强的情况下，我们的策略在各种设置中均表现出高效性和鲁棒性。