eduzhai > Applied Sciences > Engineering >

Pose2RGBD Generating Depth and RGB images from absolute positions

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 7 pages

Abstract: We propose a method at the intersection of Computer Vision and ComputerGraphics fields, which automatically generates RGBD images using neuralnetworks, based on previously seen and synchronized video, depth and posesignals. Since the models must be able to reconstruct both texture (RGB) andstructure (Depth), it creates an implicit representation of the scene, asopposed to explicit ones, such as meshes or point clouds. The process can bethought of as neural rendering, where we obtain a function f : Pose -> RGBD,which we can use to navigate through the generated scene, similarly to graphicssimulations. We introduce two new datasets, one based on synthetic data withfull ground truth information, while the other one being recorded from a droneflight in an university campus, using only video and GPS signals. Finally, wepropose a fully unsupervised method of generating datasets from videos alone,in order to train the Pose2RGBD networks. Code and datasets are available at::this https URL.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×