论文标题
无监督的3D人姿势估计的不变老师和e象学生
Invariant Teacher and Equivariant Student for Unsupervised 3D Human Pose Estimation
论文作者
论文摘要
我们提出了一种基于教师学生学习框架的新方法，用于3D人姿势估计，而无需任何3D注释或侧面信息。为了解决这个无监督的学习问题，教师网络采用基于姿势 - 词式的建模进行正规化，以估计物理上合理的3D姿势。为了处理教师网络中的分解歧义，我们提出了一个循环统一的体系结构，以促进3D旋转不变的属性来训练教师网络。为了进一步提高估计准确性，学生网络采用了新型的图形卷积网络，以灵活地直接估计3D坐标。采用了另一种促进3D旋转等值属性的循环一致的体系结构来利用几何形状的一致性，以及从教师网络中的知识蒸馏来提高姿势估计绩效。我们对人类36m和MPI-INF-3DHP进行了广泛的实验。与最先进的无监督方法相比，我们的方法将3D联合预测误差降低了11.4％，并且比使用人类360万的侧面信息的许多弱监督法。代码将在https://github.com/sjtuxcx/ites上找到。
We propose a novel method based on teacher-student learning framework for 3D human pose estimation without any 3D annotation or side information. To solve this unsupervised-learning problem, the teacher network adopts pose-dictionary-based modeling for regularization to estimate a physically plausible 3D pose. To handle the decomposition ambiguity in the teacher network, we propose a cycle-consistent architecture promoting a 3D rotation-invariant property to train the teacher network. To further improve the estimation accuracy, the student network adopts a novel graph convolution network for flexibility to directly estimate the 3D coordinates. Another cycle-consistent architecture promoting 3D rotation-equivariant property is adopted to exploit geometry consistency, together with knowledge distillation from the teacher network to improve the pose estimation performance. We conduct extensive experiments on Human3.6M and MPI-INF-3DHP. Our method reduces the 3D joint prediction error by 11.4% compared to state-of-the-art unsupervised methods and also outperforms many weakly-supervised methods that use side information on Human3.6M. Code will be available at https://github.com/sjtuxcx/ITES.
