SBIR Phase II: Semi-Automated Sports Video Search

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。小企业创新研究（SBIR）第二阶段项目的目标是将一种用于索引视频的新技术商业化。该公司的方法通过为视频生成丰富语义索引的算法自动集成语音、文本和视频信息。第一阶段的结果表明，这种方法可以被纳入一个系统，索引视频具有高精度和成本的一小部分，目前使用的方法。此外，在第一阶段的研究中，该公司已经确定了一个庞大且不断增长的消费市场（体育视频），该技术可以应用于其中。 第二阶段提案的技术目标侧重于与这些合作伙伴合作，推出基于Bluefin的初始应用程序，例如基于内容的搜索和视频丰富的梦幻体育。由于自动索引方法的准确性低，手工索引视频的成本高，这种应用目前还不可行，每月有数百万小时的新视频内容上网，满足了爆炸性的需求，重塑了互联网的性质。正如在以文本为中心的互联网的第一阶段，需要有面向文本的搜索引擎使用户能够找到他们所需要的东西一样，在迅速到来的以视频为主的互联网时代，也需要有新一代的技术来组织和有效地找到内容。蓝鳍金枪鱼实验室正在通过商业化在学术界开发的跨模态算法，开创一种新的视频组织和搜索方法。虽然这种差异化技术可以在多个目标市场中得到利用，但该公司最初的重点是体育媒体，它将为视频搜索、视频增强的梦幻体育和其他以视频为中心的应用提供独特的体验。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).The Small Business Innovation Research (SBIR) Phase II project objective is to commercialize a novel technology for indexing video. The company's approach automatically integrates information from speech, text, and video through algorithms that generate rich semantic indexes for video. The Phase I results show that this approach can be incorporated into a system that indexes video with high accuracy and at a fraction of the cost of currently used methods. Further, during the Phase I research, the company has identified a large and growing consumer market (sports video) in which the technology can be applied. The technical objectives of the Phase II proposal focus on working with such partners to roll out initial Bluefin-powered applications, such as content-based search and video-enriched fantasy sports. Such applications are currently not feasible because of the low accuracy of automated indexing methods and the high cost of manual approaches to indexing video.Millions of hours of new video content are coming online every month, feeding an exploding demand and reshaping the nature of the Internet. Just as text-oriented search engines were necessary to empower users to find what they needed during the first phase of the text-centric Internet, a new generation of technology will be necessary to organize and effectively find content in the fast-approaching video-dominated era of the Internet. Bluefin Lab is pioneering a new approach to video organization and search by commercializing cross-modal algorithms developed in Academe. While this differentiated technology can be leveraged in several target markets, the company's initial focus is on sports media where it will power a unique experience for video search, video-enhanced fantasy sports, and other video-centric applications.