ICCV Daily 2019 - Thursday

the essential matrix. What the team is doing is averaging the essential matrix before the extraction of rotation and translation. In this way, they hope to achieve more accurate results with much less time because there are less steps. He is pleased to add that they have compared their method against several state-of-the-art methods and can show that it improves the result in some cases with a faster running time. In the real world, applications include 3D reconstruction, SLAM applications, and autonomous cars . For the latter, he explains that when we want to estimate the distance to an object, we can do it with essential matrices. Using essential matrices, we can compute the relative distance between objects and the relative distance from the camera to other objects. Thinking about next steps, Amnon tells us they would like to incorporate their work into more practical SLAM applications. They want to solve the collinear case, which means that all cameras are on the same line. If you want to learn more about this work, come along to the team’s poster today at 10:30 in Hall B. 11 Amnon Geifman DA I L Y