对偶图相似性：任务迁移学习初始化选择的通用框架
Duality Diagram Similarity: a generic framework for initialization selection in task transfer learning
摘要：
在本文中，我们解决了迁移学习中的一个开放性研究问题，即在给定几个预训练模型的情况下，选择模型初始化以在新任务上实现高性能。提出了一种基于深度神经网络（DNN）之间的对偶图相似性（DDS）的高效、准确的新方法。DDS是一个通用的框架来表示和比较不同特征维度的数据。我们通过测量17个taskonomy任务的实际迁移学习性能排名与预测排名之间的对应关系，在Taskonomy数据集上验证了我们的方法。计算$17\times17$传输的基于DDS的排名需要不到2分钟，并且显示出与实际迁移学习排名的高度相关性（$0.86$），在Taskonomy基准测试中，其表现远远优于最先进的方法（$10\%$）。我们还证明了我们的模型选择方法的鲁棒性的一个新的任务，即Pascal VOC语义分割。此外，我们表明，我们的方法可以应用于选择DNN内的最佳层位置，用于NYUv2和Pascal VOC数据集上的2D，3D和语义任务的迁移学习。
英文摘要：
In this paper, we tackle an open research question in transfer learning, which is selecting a model initialization to achieve high performance on a new task, given several pre-trained models. We propose a new highly efficient and accurate approach based on duality diagram similarity (DDS) between deep neural networks (DNNs). DDS is a generic framework to represent and compare data of different feature dimensions. We validate our approach on the Taskonomy dataset by measuring the correspondence between actual transfer learning performance rankings on 17 taskonomy tasks and predicted rankings. Computing DDS based ranking for $17\times17$ transfers requires less than 2 minutes and shows a high correlation ($0.86$) with actual transfer learning rankings, outperforming state-of-the-art methods by a large margin ($10\%$) on the Taskonomy benchmark. We also demonstrate the robustness of our model selection approach to a new task, namely Pascal VOC semantic segmentation. Additionally, we show that our method can be applied to select the best layer locations within a DNN for transfer learning on 2D, 3D and semantic tasks on NYUv2 and Pascal VOC datasets.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
