论文标题
通过重构和插值的统一替代损失框架
A Unified Framework of Surrogate Loss by Refactoring and Interpolation
论文作者
论文摘要
我们介绍了Uniloss，这是一个统一的框架，旨在产生替代损失，以训练具有梯度下降的深层网络，从而减少了特定于任务的替代损失的手动设计量。我们的主要观察结果是，在许多情况下，在一批示例上评估具有性能指标的模型可以分为四个步骤：从输入到实值分数，从分数到分数对比较，从比较到二进制变量，从二进制变量到二进制变量。使用此重构，我们通过插值为每个非差异性步骤生成可区分的近似值。使用Uniloss，我们可以使用一个统一的框架为不同的任务和指标优化，与特定于任务的损失相比，实现了可比的性能。我们验证了UNILOSS对三个任务和四个数据集的有效性。代码可从https://github.com/princeton-vl/uniloss获得。
We introduce UniLoss, a unified framework to generate surrogate losses for training deep networks with gradient descent, reducing the amount of manual design of task-specific surrogate losses. Our key observation is that in many cases, evaluating a model with a performance metric on a batch of examples can be refactored into four steps: from input to real-valued scores, from scores to comparisons of pairs of scores, from comparisons to binary variables, and from binary variables to the final performance metric. Using this refactoring we generate differentiable approximations for each non-differentiable step through interpolation. Using UniLoss, we can optimize for different tasks and metrics using one unified framework, achieving comparable performance compared with task-specific losses. We validate the effectiveness of UniLoss on three tasks and four datasets. Code is available at https://github.com/princeton-vl/uniloss.
