VNect

TLDR

The study introduces the first real‑time, single‑camera method for capturing a human’s full global 3D skeletal pose with temporal consistency. The method uses a fully‑convolutional CNN pose regressor that jointly predicts 2D and 3D joint positions, followed by real‑time kinematic skeleton fitting to produce temporally stable global 3D poses without requiring cropped inputs. The approach achieves real‑time performance with accuracy comparable to state‑of‑the‑art offline monocular RGB pose estimators, outperforms RGB‑D methods in some cases, and works in diverse settings such as outdoor scenes and low‑quality commodity cameras.

Abstract

We present the first real-time method to capture the full global 3D skeletal pose of a human in a stable, temporally consistent manner using a single RGB camera. Our method combines a new convolutional neural network (CNN) based pose regressor with kinematic skeleton fitting. Our novel fully-convolutional pose formulation regresses 2D and 3D joint positions jointly in real time and does not require tightly cropped input frames. A real-time kinematic skeleton fitting method uses the CNN output to yield temporally stable 3D global pose reconstructions on the basis of a coherent kinematic skeleton. This makes our approach the first monocular RGB method usable in real-time applications such as 3D character control---thus far, the only monocular methods for such applications employed specialized RGB-D cameras. Our method's accuracy is quantitatively on par with the best offline 3D monocular RGB pose estimation methods. Our results are qualitatively comparable to, and sometimes better than, results from monocular RGB-D approaches, such as the Kinect. However, we show that our approach is more broadly applicable than RGB-D solutions, i.e., it works for outdoor scenes, community videos, and low quality commodity RGB cameras.