arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1910.09447cs.CV
收藏

用校准的字体改善风格迁移

Improving Style Transfer with Calibrated Metrics

University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)

Mao-Chuang Yeh, Shuai Tang, Anand Bhattad, Chuhang Zou, David Forsyth

摘要:

风格转移方法产生转移图像,该转移图像是以风格图像的方式呈现的内容图像。我们试图了解如何改善风格转移。要做到这一点,就需要有定量评价程序,但目前的评价是定性的,主要涉及用户研究。我们描述了一种新的定量评价程序。我们的过程依赖于两个统计量:有效性(E)统计量测量给定样式已被转移到目标的程度,一致性(C)统计量测量原始图像内容被保留的程度。我们的统计数据被校准到人类偏好:具有较大E值(resp C)的目标将可靠地被人类受试者在风格(resp.内容)。我们使用这些统计数据来研究一些神经风格转移(NST)方法的相对性能,揭示了一些有趣的特性。可接受的方法位于帕累托边界(即改进E会减少C,反之亦然)。有三种方法是可以接受的:通用风格转换产生非常好的C,但弱E;修改用于Gatys损失的优化产生具有强E和强C的方法;修改的跨层方法在C的强成本下具有稍好的E。虽然直方图损失改进了Gatys方法的E统计量,但它并没有使该方法变得可接受。令人惊讶的是,风格权重在提高EC分数方面的作用相对较小,并且迁移中的大多数变化都是由风格本身解释的(这意味着实验者可能会被选择的风格误导)。

英文摘要:

Style transfer methods produce a transferred image which is a rendering of a content image in the manner of a style image. We seek to understand how to improve style transfer. To do so requires quantitative evaluation procedures, but the current evaluation is qualitative, mostly involving user studies. We describe a novel quantitative evaluation procedure. Our procedure relies on two statistics: the Effectiveness (E) statistic measures the extent that a given style has been transferred to the target, and the Coherence (C) statistic measures the extent to which the original image's content is preserved. Our statistics are calibrated to human preference: targets with larger values of E (resp C) will reliably be preferred by human subjects in comparisons of style (resp. content). We use these statistics to investigate the relative performance of a number of Neural Style Transfer(NST) methods, revealing several intriguing properties. Admissible methods lie on a Pareto frontier (i.e. improving E reduces C or vice versa). Three methods are admissible: Universal style transfer produces very good C but weak E; modifying the optimization used for Gatys' loss produces a method with strong E and strong C; and a modified cross-layer method has slightly better E at strong cost in C. While the histogram loss improves the E statistics of Gatys' method, it does not make the method admissible. Surprisingly, style weights have relatively little effect in improving EC scores, and most variability in the transfer is explained by the style itself (meaning experimenters can be misguided by selecting styles).

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。

↑