We propose to identify directions invariant to a given classifier so that these directions can be controlled in tasks such as style transfer. While orthogonal decomposition is directly identifiable when the given classifier is linear, we formally define a notion of orthogonality in the non-linear case. We also provide a surprisingly simple method for constructing the orthogonal classifier (a classifier utilizing directions other than those of the given classifier). Empirically, we present three use cases where controlling orthogonal variation is important: style transfer, domain adaptation, and fairness. The orthogonal classifier enables desired style transfer when domains vary in multiple aspects, improves domain adaptation with label shifts and mitigates the unfairness as a predictor. The code is available at http://github.com/Newbeeer/orthogonal_classifier
翻译:我们建议确定给给定分类师的变量方向, 以便在样式传输等任务中控制这些方向。 虽然当给定分类师为线性时, 正方形分解可直接识别, 我们正式定义非线性案例的正方形分解概念。 我们还为构造正方形分类师( 使用给定分类师以外方向的分类师) 提供了一个令人惊讶的简单方法 。 随机性, 我们使用三种情况来控制正方形变异很重要: 样式转移、 域适应、 公平性 。 当域在多个方面不同时, 正方形分类师可以实现理想的样式转移, 改进标签变化的域适应, 并减轻作为预测器的不公平性 。 代码可在 http://github. com/ Newbeeeer/ orthogonal_ claslicer 上查阅 。