arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2008.11932cs.CV
收藏

基于布局的属性引导图像生成

Attribute-guided image generation from layout

University of British Columbia(不列颠哥伦比亚大学);Bank of Montreal(蒙特利尔银行)

Ke Ma, Bo Zhao, Leonid Sigal

摘要:

最近的方法在从结构化输入生成图像方面取得了巨大成功,语义分割、场景图或布局。虽然这些方法允许在图像级指定对象及其位置,但它们缺乏在实例级指定这些对象的视觉外观的保真度和语义控制。为了解决这个问题,我们提出了一种新的图像生成方法,使实例级属性控制。具体来说,我们的属性引导生成模型的输入是一个元组,它包含:(1)对象边界框,(2)对象类别和(3)每个对象的(可选)属性集。输出是生成的图像,其中所请求的对象位于所需位置并具有规定的属性。几个损失协同工作,以鼓励准确,一致和多样化的图像生成。在Visual Genome数据集上的实验表明,该模型能够有效地控制生成图像中的对象级属性,并验证了从布局任务生成图像时对象-属性分离表示的可行性。此外,从我们的模型生成的图像具有更高的分辨率,对象分类精度和一致性,与以前的最先进的。

英文摘要:

Recent approaches have achieved great success in image generation from structured inputs, e.g., semantic segmentation, scene graph or layout. Although these methods allow specification of objects and their locations at image-level, they lack the fidelity and semantic control to specify visual appearance of these objects at an instance-level. To address this limitation, we propose a new image generation method that enables instance-level attribute control. Specifically, the input to our attribute-guided generative model is a tuple that contains: (1) object bounding boxes, (2) object categories and (3) an (optional) set of attributes for each object. The output is a generated image where the requested objects are in the desired locations and have prescribed attributes. Several losses work collaboratively to encourage accurate, consistent and diverse image generation. Experiments on Visual Genome dataset demonstrate our model's capacity to control object-level attributes in generated images, and validate plausibility of disentangled object-attribute representation in the image generation from layout task. Also, the generated images from our model have higher resolution, object classification accuracy and consistency, as compared to the previous state-of-the-art.

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。

↑