On the Role of Depth Predictions for 3D Human Pose Estimation

Abstract

AbstractFollowing the successful application of deep convolutional neural networks to 2D human pose estimation, the next logical problem to solve is static 3D human pose estimation from monocular images. While previous solutions have shown some success, they do not fully utilize the depth information from the 2D inputs. With the goal of addressing this depth ambiguity, we build a system that takes 2D joint locations as input along with their estimated depth value and predicts their 3D positions in camera coordinates. Our system out performs comparable frame-by-frame 3D human pose estimation networks on the largest publicly available 3d motion data set, Human 3.6M. To provide further evidence for the usefulness of predicted depth values in the 3D pose estimation problem, we perform an extensive statistical analysis showing that even with potentially noisy depth predictions there is still a statistically significant correlation between the predicted depth value and the true depth value.KeywordsConvolutional neural networkPose estimation3DDepthMonocular imagesMachine learning