康奈尔大学本科生视觉与学习（CUVL）研究获Facebook资助！-新东方前途出国

　　Cornell Undergraduate Vision and Learning (CUVL) Research Recognized by a Grant from Facebook

　　Research Areas: Artificial IntelligenceGraphicsScientific ComputingRelated People: Serge Belongie Kavita Bala Bharath Hariharan 

　　研究领域：人工智能图形科学计算相关人员：Serge Belongie Kavita Bala Bharath Hariharan

　　Cornell CS undergraduates Qian Huang, Isay Katsman, Zeqi Gu, and Horace He—members of the Cornell Undergraduate Vision and Learning (CUVL) club—under the direction of CS Professor Serge Belongie and Ser-Nam Lim (of Facebook AI), presented their research "Enhacing Adversarial Example Transferability with an Intermediate Level Attack" at the International Conference in Computer Vision (ICCV 2019) in Seoul, South Korea, as reported earlier. 

　　康奈尔大学CS本科生Qian Huang、Isay Katsman、Zeqi Gu和Horace He 都是康奈尔大学本科生视觉与学习（CUVL）俱乐部成员，在CS教授Serge Belongie和Ser Nam Lim（Facebook AI）的指导下，如前所述，他们在韩国首尔举行的计算机视觉国际会议（ICCV2019）上提出了“通过中级攻击增强对抗性示例可转移性”的研究。

　　Subsequent to the presentation of research that benefitted directly from Facebook support, the tech giant feted the one-year old group with another substantial grant. Read more about their successes in the Cornell Chronicle story “CS undergrads’ research sets sights on image hackers.” As reported by Melanie Lefkowitz: 

　　在介绍了直接受益于Facebook支持的研究成果后，这家科技巨头又为这家成立一年的集团提供了一笔可观的资助。在《康奈尔纪事报》（Cornell Chronicle）上阅读更多关于他们成功的文章“CS本科生的研究着眼于图像黑客”。如梅兰妮·莱夫科维茨（Melanie Lefkowitz）所报道： 

　　CS Professor Serge Belongie is working with Facebook on a project to combat “deepfakes”—faked audio and video created using artificial intelligence. 

　　CS教授Serge Belongie正在与Facebook合作一个项目，以打击使用人工智能创建的“deepfakes”伪造音频和视频。 

　　The Cornell Undergraduate Vision and Learning (CUVL) club used the money to purchase graphics processing units—costly, powerful processors that are necessary for training most machine learning models, and generally available only to graduate students and faculty. Using central processing units—which is how most computers function—the thousands or millions of iterations needed to train machine learning algorithms could take weeks or months. 

　　康奈尔大学本科生视觉与学习（CUVL）俱乐部用这笔钱购买了图形处理单元，成本高昂，功能强大，是训练大多数机器学习模型所必需的处理器，通常只提供给研究生和教员。使用中央处理单元（这是大多数计算机的工作方式）训练机器学习算法所需的数千或数百万次迭代可能需要数周或数月的时间。 

　　“If you have a good idea but you can’t verify it because it’s going to take a year to do the computation, then you can’t really do the research,” Huang said. “So now we have the tools we need to do the work.” 

　　“如果你有一个好主意，但由于要花一年的时间来计算而无法验证，那么你就不能真正地进行研究，”黄说。“因此，现在我们有了完成工作所需的工具。”

　　Facebook has since donated another [round of funding] to the club, which will pay for eight more graphics processing units, the students said. 

　　学生们说，Facebook此后又向俱乐部捐赠了（一轮资金），俱乐部将再支付8个图形处理单元的费用。 

　　In their paper, the students tackled the problem of adversarial examples—tiny tweaks to an image that are undetectable to the human eye but completely confusing to a neural network tasked with classifying images. Adversarial examples created by hackers or others with malicious intent could potentially disorient autonomous cars, for example, or subvert image recognition. 

　　在他们的论文中，学生们解决了对抗性例子的问题，即对人眼看不到的图像进行微小的调整，但对一个负责对图像进行分类的神经网络来说完全是一种混乱。由黑客或其他具有恶意意图的人创建的敌对示例可能会使自动驾驶汽车失去方向感，或者破坏图像识别。

　　如需进一步了解，或有任何相关疑问，欢迎在线咨询留学专家。如果您对自己是否适合留学还有疑虑，欢迎参与前途出国免费评估，以便给您进行准确定位。