This work introduces Conditional Image Retrieval (CIR) systems: IR methods that can efficiently specialize to specific subsets of images on the fly. These systems broaden the class of queries IR systems support, and eliminate the need for expensive re-fitting to specific subsets of data. Specifically, we adapt tree-based K-Nearest Neighbor (KNN) data-structures to the conditional setting by introducing additional inverted-index data-structures. This speeds conditional queries and does not slow queries without conditioning. We present two new datasets for evaluating the performance of CIR systems and evaluate a variety of design choices. As a motivating application, we present an algorithm that can explore shared semantic content between works of art of vastly different media and cultural origin. Finally, we demonstrate that CIR data-structures can identify Generative Adversarial Network (GAN) "blind spots": areas where GANs fail to properly model the true data distribution.

中文翻译：

---

条件图像检索

这项工作介绍了条件图像检索 (CIR) 系统：IR 方法可以有效地专门用于动态图像的特定子集。这些系统拓宽了 IR 系统支持的查询类别，并消除了对特定数据子集进行昂贵的重新拟合的需要。具体来说，我们通过引入额外的倒排索引数据结构，使基于树的 K-最近邻 (KNN) 数据结构适应条件设置。这加快了条件查询，并且不会在没有条件的情况下减慢查询。我们提供了两个新的数据集，用于评估 CIR 系统的性能并评估各种设计选择。作为一个激励应用程序，我们提出了一种算法，可以探索截然不同的媒体和文化起源的艺术作品之间的共享语义内容。最后，