一秒拯救名画，这种人工AI修复技术太火了-天涯论坛

m5k1umn 发表于 2024-8-19 05:39:59

一秒拯救名画，这种人工AI修复技术太火了

<div style="color: black; text-align: left; margin-bottom: 10px;">
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/2d6a19602cbf465a8902e7199e86e95d~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=hEC2pq7zd5NGjdVekH%2FDTB0zfuo%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/941a21442ca84238a0d634e90e8e7a3e~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=YFUF1aIq3nl8zuJt1z6ETXd7i08%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/c7fdc849540146178802fda45b46021c~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=PmCHqsm1lp9l188kOyWzxQ1WNnk%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/d1bed688a51f4d4eafded00c6f511493~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=MSfO8NCHWY%2BHM6%2FHyFKa58yjQnw%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/51081bd1534d47898bccd2f5f46f74ce~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=Jsh6YllWqxM1iET7lHGc6sNCJac%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/4c261c84463f480f8ccd5de4bdd60135~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=XPaMuB%2Fq5hQB%2FX7XzgiZF1BlfSU%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/5eea08e928f14253aa752fa0b0929b9f~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=iexv1u6d%2FBHEYpyJOrVFLAQ5yDE%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/1f8e6538417c419ebd8a0134b0818501~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=PWlofRp9D3MD371lI%2FACsrrKtE0%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/87d2cfefad1344739309de1b4d23c923~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=lIwijSGx5BT6AdQRSSUIUt0lkMw%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/6b13500b6bb84344b334496cd628017f~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=dES28%2F5oPUExx4IULTwURzw%2Bgao%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/7f74a44b44ef4e218f71d153ab2d6373~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=Dq8%2BnI%2FEnRN%2Ff9Gwuwk5ZaB%2BUXE%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/da6d88bd36674b8495f8c567b2c00a15~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=6DJ4zT8WmEYqHlyia%2BAfyLxvxnI%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/04f013ada43a4567bf903c99213e2892~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=1UA91nMCIbPQAvjn07IRFnePLUc%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/8a7c280cb0d941e38cf59bfe55d0210f~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=jB1xeZ4mZ0%2B1bpWeOqBHDOqxph8%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/24146d9b779c40ef8ff8b9b8e43230e7~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=ufnhNrnC%2FNYL5Vy%2Flh%2Bh9SI7koQ%3D" style="width: 50%; margin-bottom: 20px;"></div>
 X-ray images of polyptych wings, or other artworks painted on both sides of their support, contain in one image content from both paintings, making them difficult for experts to “read.” To improve the utility of these x-ray images in studying these artworks, it is desirable to separate the content into two images, each pertaining to only one side. This is a difficult task for which previous approaches have been only partially successful. Deep neural network algorithms have recently achieved remarkable progress in a wide range of image analysis and other challenging tasks. We, therefore, propose a new self-supervised approach to this x-ray separation, leveraging an available convolutional neural network architecture; results obtained for details from the Adam and Eve panels of the Ghent Altarpiece spectacularly improve on previous attempts.
 多联画屏的X射线扫描图，或双面绘制的其他艺术作品，一般将正反两面两幅作品中的意象集成到一幅图像中，这无疑会加大专家“解析”画作的难度。为了让X射线技术在解析画作的过程中发挥更大效用，科研有意将这种合二为一的画像内容拆分为两幅图像，每幅图像都各自表率其中一面的意象。这项任务非常艰巨，传统办法只能取得事倍功半的效果。近期，深度神经网络算法已在图像分析等难度系数较高的行业中得到广泛认可。因此呢，科研提出了一种新型自我监督机制的X射线分离办法，办法采用切实可行的卷积神经网络框架；以上办法在《根特祭坛画》的亚当与夏娃组画上取得了明显效果。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/8243232309564293807b2c710a116eac~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=8G%2B%2BBJ6ZjEYxkwAV4NLC6QXRJvI%3D" style="width: 50%; margin-bottom: 20px;"></div>
 In the art investigation domain, increasing use of extremely high-resolution digital imaging techniques is being made in parallel with the widespread adoption of a range of recent imaging and analytical modalities not previously applied in the field (e.g., hyperspectral imaging, macro x-ray fluorescence scanning, and novel forms of imaging x-ray radiography). These techniques mean that there is a wealth of digital data available within the sector, offering huge scope to provide new insights but also presenting new computational challenges to the domain.
 就艺术科研行业而言，高分辨率数字影像技术日前已得到越来越多的应用，而影像和分析模式，虽然之前未被采纳应用（如在高光谱影像、宏观X射线荧光扫描、新奇的X射线影像照相术行业），但近期亦被喜爱有加，这两种技术在艺术科研行业并驾齐驱。这些技术的广泛应用，说明艺术科研行业仍有海量的图像数据资源可被挖掘，这些技术为探究更加多新的发掘供给了一个宏观的视角，但亦向艺术行业提出了新的计算技术挑战。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/b7b77b6c25c7423e978649ec2183d2d3~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=Lszrkx%2BQbi%2BTsm%2F5Y9Vh3%2Fbc%2FxI%3D" style="width: 50%; margin-bottom: 20px;"></div>
 In the past decades, various other disciplines, experiencing similar data growth, have benefited greatly from recent breakthroughs in artificial intelligence. The availability of cutting-edge machine learning algorithms, as well as the enhanced computation power and frameworks necessary to deal with massive datasets, have yielded outstanding results in computer vision, speech recognition, speech translation, natural language processing, and more.
 在过去的几十年里，其他各个行业亦见证了这一数据趋势，并从人工智能取得的各个技术突破中受益匪浅。尖端设备学习算法的应用以及在处理海量数据集中发挥的强大计算能力和框架，在计算机图像、语音识别、语音翻译、自然语言处理等多个行业取得了明显成果。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/93f1b31badbc48bf9dc30f51aff26543~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=0SQZatrYdbJjhhn%2Bv0Z6YoV%2Bo1c%3D" style="width: 50%; margin-bottom: 20px;"></div>
 It is therefore natural to develop similar techniques to address challenging tasks arising in art investigation, including material identification within different strata of paintings, analysis of brush stroke style, automatic canvas thread counting, digital inpainting of cracks, and visualization of concealed designs and underdrawings.
 因此呢，研发关联技术并利用这些技术来处理艺术科研中遇到的困难，这一点毋庸置疑，这些技术包含区别层次的绘画材料识别、笔触风格分析、自动画布线程计数、裂缝的数字生成、隐匿设计和底图的可视化展示。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/4f24b8a132504f4da833fd9ec6942aef~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=mhqmRVlGQAClS3jaApa8J8HY550%3D" style="width: 50%; margin-bottom: 20px;"></div>
 This paper deals with a challenging image processing task arising in the context of the painting The Adoration of the Mystic Lamb, painted in 1432 by the brothers Hubert and Jan Van Eyck and more commonly known as the Ghent Altarpiece, shown in Fig. 1. This piece is one of the most admired and influential paintings in the history of art, showcasing the Van Eyck brothers’ unrivaled mastery of the oil painting technique. Over the centuries, the monumental polyptych (350 cm by 470 cm when open) has been prized for its stunning rendering of different materials and textures and its complex iconography. Originally, the polyptych consisted of four central panels, and then two wings each consisting of four panels painted on both sides so that entirely different sets of images and iconography could be seen depending on whether or not the wings were open (for example on Feast days). This paper focuses on two of the double-sided panels, depicting Adam and Eve on the interiors (and with the Annunciation and interior scenes on the outside).
 本文旨在科研《奥秘羔羊之爱》作品中繁杂的图像处理技术，《奥秘羔羊之爱》由胡伯特·凡· 艾克和扬·凡· 艾克兄弟在 1432 年创作，众所周知为《根特祭坛画》之名，如图1所示。这幅作品是艺术史上备受赞誉和最有影响力的画作之一，展示了凡·艾克兄弟不可超越的油画技巧。因其绝妙的材料运用、纹理渲染和纷繁繁杂的图像技术，这幅蔚为壮观的联画作品（完全打开时长 3.5 米，宽 4.7 米）在几个世纪败兴始终备受赞誉。最初这幅联画作品由四个中心镶板构成，镶板两侧的每一个侧翼由四个面板构成，面板为双面绘制，这般一来，侧翼的打开与否（例如节日会打开侧翼）便决定了所展示的区别图像和肖像画作。本论文重点科研双面作画的面板，对内页的《亚当与夏娃》（和外边翼板上的《天使报喜》以及两边外侧上的内页画作）进行说明。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/1abe9b2f024f43bc8a1aec61de088919~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=v978VOehew%2FUWP20%2FG1hiUk5MHg%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/062ee053d56742d292f859904a287d18~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=J49rS45KVJBb3CPL42wXXBkbSBw%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/90f18f4d5bac4e41a525595931451251~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=S1xp7jBONPHnVZc%2BTO6nPHnmwJA%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/80b896da396f41a08afa9a8d2e6a6b40~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=dUL5L7IXySIVFQfH%2FsuiSx3IowQ%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/20ee53fbc873443c924fdab5f29be98d~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=jCcNpoVil8S9MbG4WZpBCE%2FVqp0%3D" style="width: 50%; margin-bottom: 20px;"></div>
 《根特祭坛画》
 做为15世纪尼德兰美术的标志,《根特祭坛画》这幅出此刻尼德兰文艺复兴初期的艺术巨制，拥有一种里程碑的道理。题材虽源自于宗教，但画家以对现实世界的肯定和赞美的态度，以及对名人的细致和写实的描绘，构成为了这幅作品的基调，从而使全部画面充满诗意，并具有没有穷艺术魅力。《根特祭坛画》能够叫作为世界上第1件真正的油画作品，色彩鲜明，辉煌艳丽，经过数百年之后，画面仍然如初，这在当时的确是一种绘画技法上的突进。因此，在绘画史上的道理，远远超出了通常的思想内容和艺术形式处理上的革新和独创，而是开创了全部欧洲绘画的新纪元。
 《根特祭坛画》是一种多翼式“开闭形”祭坛组画。这幅祭坛画，乃置放在教堂圣坛的前面，木板构成，由内外共20个画面形成一种折叠式画障。每逢节日的礼拜盛会，祭坛的两翼伴同着音乐旋律被徐徐打开，辉煌夺目的12幅祭坛画便呈此刻人们的眼前。
 凡·艾克兄弟指的是14世纪到15世纪尼德兰画家胡伯特·凡·艾克（1370—1426年）和扬·凡·艾克（1385/90—1441年）兄弟。1415年根特市长约多库斯·威德向胡伯特·凡·艾克订制祭坛画，他画了10年，未完成就去逝了，后由弟弟扬·凡·艾克继续完成，安顿在根特圣贝文教堂。这是一组拥有里程碑道理的划时代巨作，它标志着一个新时代的到来和人文主义艺术的诞生，并奠定了尼德兰文艺复兴艺术的基本。因为《根特祭坛画》是运用油调颜色绘制而成，凡·艾克兄弟作为欧洲油画的创始人。
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/50fbdaafe93348df8e8e34c06ceb94cd~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=rP9gkmih2yBlKN9JmsCbLTTEy8A%3D" style="width: 50%; margin-bottom: 20px;"></div>
 AI助力艺术行业
 虽然这些AI都有十八班武艺，但其实它们都基于图像识别技术。当然，最重要的是，它们的研发都用到了Google的TensorFlow——日前最主流的设备学习研发框架。
 经过TensorFlow，Google艺术与文化研发了海量的艺术项目，例如艺术自拍（Art Selfie），拍一张自拍，你就能找到自己在艺术中的“分身”。
 得益于艺术作品的数字化尝试，这些想象力丰富的项目才可实现，并让更加多艺术触及到公众。然而，世界上还有海量艺术和文化遗产无实现数字化，巴西国家博物馆便是一个令人惋惜的例子。上千万件藏品遭遇大火，但因为该博物馆的数字化非常落后，咱们只能和非常多藏品告别了。
 好在，AI在艺术和文化行业的发展，已然揭示出数字化的必要性。倘若咱们人类本身无能力让隐匿的艺术重见天日，AI或许会是一个得力的助手。而针对艺术家和设计师来讲，AI乃至能够帮忙激发创造力。Google此前推出的AutoDraw，就期盼设计师或画家在落笔后，AI能给出有用的绘画意见。
 虽然AI已然能做这么多了，但亦诱发出一个有趣的问题：AI真的能创造出艺术吗？这个问题很难回答，一方面，AI必须依赖海量的训练数据，它的创作其实还是按照人类已有的作品实现的。另一方面，创作艺术必然要理解艺术，这背面触及到非常多深层次的问题，例如设备能否有认识，审美体验又到底是什么。
 你怎么看？欢迎大众文末给阿伦老师留言！
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/831ac92d3c4947b3ad5d810f67513edb~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=1jidD3%2B6jvA9a6Mm8CePdBQBZ74%3D" style="width: 50%; margin-bottom: 20px;"></div>
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/c32a56afba6d485685881095bdccd9f6~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=X9b6eeiDQamByCQsxFZSTOVpPzE%3D" style="width: 50%; margin-bottom: 20px;"></div>
 全世界100多个国家的科研叫作：女性权利越高，百姓更健康
 多亏了古丝绸之路，咱们才可吃上美味的苹果
 盐竟然是过敏免疫反应罪魁祸首之一？
 揭秘喵星人为么爱吃草？
 认识状态面面谈
 适度财富不平等或能促进群身体合作
 秋天有多美？一品这11句诗文便知
 奥巴马投资制作的首部电影：中国老板投资拯救美国工厂, 诱发全世界轰动
 每日学点社会学：语言与认知
 每日学点社会学：社会的类型
 每日学点心理学，人是怎样感知的？
 <div style="color: black; text-align: left; margin-bottom: 10px;"><img src="https://p3-sign.toutiaoimg.com/pgc-image/f5cad0b113a24dd0b9dd1fb4a84f479e~noop.image?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1723956305&x-signature=eeliy7v3uthlhhkckU9AErzGRPE%3D" style="width: 50%; margin-bottom: 20px;"></div>
 ■ 托福提分月计划课程（在线小班）
 ■ SAT精英课程（在线小班）
 ■ 托福/SAT/GRE写作批改服务（48h内回复）
 *更多TOEFL丨SAT丨AP丨GRE实用的提分秘籍和有趣好玩的视频课程，关注微X公众号"SATPEDIA"。
</div>

qzmjef 发表于 2024-10-1 12:34:43

回顾历史，我们感慨万千；放眼未来，我们信心百倍。

4lqedz 发表于昨天 03:41

你的见解独到，让我受益匪浅，非常感谢。

页: [1]

天涯论坛's Archiver

一秒拯救名画，这种人工AI修复技术太火了