A good example that we live in the era of Big Data is that, as we’ve moved from super-8-film home movies to ever-present smartphones, we’ve all begun to generate so much visual imagery that we seldom look at a given video more than once. Worse, when we do want to find a video clip, it’s lost among thousands of others. Machine intelligence researchers Shoou-I Yu and Lu Jiang, working with colleagues on Carnegie Mellon University’s Alexander Hauptmann’s Informedia project and at PSC have developed E-Lamp, a system of “event detectors” designed to search for events in videos without human intervention. Such a detector could help us all keep better tabs of our videoelectronic lives. The task of finding a video of a birthday party, for example, is fairly easy for a person. But it’s extremely hard for a computer: All the cues a machine might use to spot a video, including color, shape, sounds and even captions can be misleading. To read more and watch the video, please visit http://www.psc.edu/index.php/teaching-the-machine.