DeepTAM: Deep Tracking and Mapping [chapter]

Huizhong Zhou, Benjamin Ummenhofer, Thomas Brox
2018 Lecture Notes in Computer Science  
We present a system for keyframe-based dense camera tracking and depth map estimation that is entirely learned. For tracking, we estimate small pose increments between the current camera image and a synthetic viewpoint. This significantly simplifies the learning problem and alleviates the dataset bias for camera motions. Further, we show that generating a large number of pose hypotheses leads to more accurate predictions. For mapping, we accumulate information in a cost volume centered at the
more » ... rrent depth estimate. The mapping network then combines the cost volume and the keyframe image to update the depth prediction, thereby effectively making use of depth measurements and image-based priors. Our approach yields state-of-the-art results with few images and is robust with respect to noisy camera poses. We demonstrate that the performance of our 6 DOF tracking competes with RGB-D tracking algorithms.We compare favorably against strong classic and deep learning powered dense depth algorithms.
doi:10.1007/978-3-030-01270-0_50 fatcat:a4mjbsavszd7tlf3nszogkvcty