SGER: Hyper-Resolution Rendering and Display

在这个提议中，我们简要地描述了高级图像原语和自适应无帧采样，以及我们在每个方面的初步结果。然后，我们列出了我们的研究计划，在我们的主要和次要应用领域：渲染和显示的背景下，调查这两种以变化为中心的图像策略。在这两个领域，我们的重点是实现超分辨率图像的技术，但我们也认为上面列出的其他好处（更准确，更便携，更灵活）将从我们的努力中出现。我们的目标雄心勃勃，但很有基础：我们并不打算真正构建新的渲染硬件和显示器，而是奠定必要的智力基础。我们还与业界合作伙伴积极合作，有能力将这些想法付诸实践，包括微软研究院、英特尔和三菱电机研究实验室。智力价值：这项研究将涉及对计算机图像的根本性反思，为图像处理、计算机视觉、机器人、计算机图形学和图像显示等成熟领域的研究开辟新的途径。此外，我们的超分辨率渲染技术使十亿像素的显示器在许多其他科学领域具有巨大的潜力。在人机交互中，许多人认为这样的显示是未来的平台，而我们的工作将使这个平台的研究成为可能。可视化研究人员还将能够研究十亿像素显示器的明显潜力，以提高对来自各种领域的复杂数据的理解。更广泛的影响：微软和三菱确信大规模和高分辨率图像将成为消费技术市场的重要组成部分。我们同意并瞄准家庭和办公室的交互式、打印机分辨率的渲染和显示。我们相信这种技术可以改变商业和个人成像行业。此外，这种渲染的带宽改进也可以很容易地应用于手机和pda上的移动图形。我们与微软、三菱和英特尔的积极研究关系将使我们脚踏实地，确保我们的想法对世界产生影响。我们提出的研究也将对教育产生直接影响。我们计划开设一门计算机摄影课程。我们在教育合作方面也有着悠久的历史，包括与博物馆的合作以及与城市设计师和建筑师的指导。超分辨率显示及其提高的精度和可扩展性将使这些努力受益，并带来新的机遇。

In this proposal we briefly describe high-level image primitives and adaptive frameless sampling in moredetail, along with our preliminary results in each. We then lay out our research plans for investigatingthese two strategies for change-focused imagery in the context of our primary and secondary applicationareas: rendering, and display. In both areas our focus is on enabling technology for hyper-resolutionimagery, but we also argue that the other benefits listed above (more accurate, more portable, and moreflexible) will emerge from our efforts. Our goals are ambitious but grounded: we do not propose toactually build new rendering hardware and monitors, but instead to lay the necessary intellectualgroundwork. We also have active collaborations with partners in industry capable of taking these ideas topractice, including Microsoft Research, Intel, and Mitsubishi Electric Research Labs.Intellectual merit: This research will involve a fundamental rethinking of computer imagery, openingnew avenues of research in well-established fields such as image processing, computer vision, robotics,computer graphics, and image display. Furthermore, the gigapixel displays enabled by our hyperresolutionrendering has great potential as a tool in many other scientific fields. In human-computer interaction, many view such displays as the platform of the future, and our work will make research on this platform possible. Visualization researchers will also be able to study the obvious potential of gigapixel displays for improving understanding of complex data from a wide variety of fields.Broader impacts: Microsoft and Mitsubishi are convinced that large-scale and high resolution imagerywill be an important part of the consumer technology market. We agree and are targeting interactive,printer-resolution rendering and display in the home and office. We believe such technology couldtransform the business and personal imaging industries. Moreover, the bandwidth improvements for suchrendering might also be easily applied to mobile graphics on phones and PDAs. Our active researchrelationships with Microsoft, Mitsubishi and Intel will keep us grounded and ensure that our ideas havereal-world impact. Our proposed research will also have direct educational impact. We plan to develop acomputational photography course. We also have a strong history of educational collaborations, includingoutreach with museums as well as instruction with urban designers and architects. Hyper-resolutiondisplay, and the improved accuracy and scalability it promises, will benefit these efforts and enable newopportunities.