3D human pose estimation in video with temporal convolutions and semi-supervised training

11/28/2018
by   Dario Pavllo, et al.
0

In this work, we demonstrate that 3D poses in video can be effectively estimated with a fully convolutional model based on dilated temporal convolutions over 2D keypoints. We also introduce back-projection, a simple and effective semi-supervised training method that leverages unlabeled video data. We start with predicted 2D keypoints for unlabeled video, then estimate 3D poses and finally back-project to the input 2D keypoints. In the supervised setting, our fully-convolutional model outperforms the previous best result from the literature by 6 mm mean per-joint position error on Human3.6M, corresponding to an error reduction of 11 significant improvements on HumanEva-I. Moreover, experiments with back-projection show that it comfortably outperforms previous state-of-the-art results in semi-supervised settings where labeled data is scarce. Code and models are available at https://github.com/facebookresearch/VideoPose3D

READ FULL TEXT
research
02/22/2020

Back to the Future: Joint Aware Temporal Deep Learning 3D Human Pose Estimation

We propose a new deep learning network that introduces a deeper CNN chan...
research
03/08/2023

Semi-Supervised 2D Human Pose Estimation Driven by Position Inconsistency Pseudo Label Correction Module

In this paper, we delve into semi-supervised 2D human pose estimation. T...
research
06/10/2021

Adversarial Motion Modelling helps Semi-supervised Hand Pose Estimation

Hand pose estimation is difficult due to different environmental conditi...
research
05/14/2021

TriPose: A Weakly-Supervised 3D Human Pose Estimation via Triangulation from Video

Estimating 3D human poses from video is a challenging problem. The lack ...
research
07/14/2022

Semi-Supervised Temporal Action Detection with Proposal-Free Masking

Existing temporal action detection (TAD) methods rely on a large number ...
research
09/02/2020

Monocular 3D Detection with Geometric Constraints Embedding and Semi-supervised Training

In this work, we propose a novel single-shot and keypoints-based framewo...

Please sign up or login with your details

Forgot password? Click here to reset