Generative Adversarial Networks can produce images of remarkable complexity and realism but are generally structured to sample from a single latent source ignoring the explicit spatial interaction between multiple entities that could be present in a scene. Capturing such complex interactions between different objects in the world, including their relative scaling, spatial layout, occlusion, or viewpoint transformation is a challenging problem. In this work, we propose a novel self-consistent Composition-by-Decomposition network to compose a pair of objects. Given object images from two distinct distributions, our model can generate a realistic composite image from their joint distribution following the texture and shape of the input objects. We evaluate our approach through qualitative experiments and user evaluations. Our results indicate that the learned model captures potential interactions between the two object domains, and generates realistic composed scenes at test time.

中文翻译：

---

Compositional GAN：学习图像条件二元组合

生成对抗网络可以生成具有非凡复杂性和真实感的图像，但通常结构化为从单个潜在源中采样，忽略场景中可能存在的多个实体之间的显式空间交互。捕捉世界上不同对象之间如此复杂的交互，包括它们的相对缩放、空间布局、遮挡或视点变换，是一个具有挑战性的问题。在这项工作中，我们提出了一种新颖的自洽组合分解网络来组合一对对象。给定来自两个不同分布的对象图像，我们的模型可以根据输入对象的纹理和形状从它们的联合分布生成逼真的合成图像。我们通过定性实验和用户评估来评估我们的方法。