By Evan Ackerman
This little web app does a fairly convincing job of turning one single 2D picture into a 3D scene that you can fly around with your keyboard. Since a single 2D picture doesn’t contain much in the way of depth information, researchers from Stanford University had to get creative to extract a depth map from an image:
“To convert the still images into 3-D visualizations, Andrew Ng, an assistant professor of computer science, and Ashutosh Saxena, a doctoral student in computer science, developed a machine-learning algorithm that associates visual cues, such as color, texture, and size, with certain depth values based on what they have learned from studying two-dimensional photos paired with 3-D data. For example, says Ng, grass has a distinctive texture that makes it look very different close up than it does from far away. The algorithm learns that the progressive change in texture gives clues to the distance of a patch of grass.”
Sure, this is nowhere near as slick as Photosynth, but you can do it with any of your pictures without any additional information. It’s free; just register on the website and give it a try. You can upload any JPEG <5mb, and after some processing (it took about 5 minutes for me at 3am PST), you can fly around your image on the Make3D website. Since it's using purely visual cues, the algorithm is far from perfect, and it works best on landscapes. Best thing you can do is be patient and try a few different images... When it gets it right, it's very cool. [ Make3D ] VIA [ Technology Review ]
