Procrustean Regression Networks: Learning 3D Structure of Non-Rigid Objects from 2D Annotations

TL;DR

This paper introduces Procrustean Regression Networks, a novel deep learning framework that learns 3D structures of non-rigid objects from 2D annotations by automatically estimating rotations, outperforming existing methods on multiple datasets.

Contribution

The paper presents a new loss function that automatically determines suitable rotations, enabling effective 3D reconstruction from 2D data without complex rotation regression.

Findings

Outperforms state-of-the-art on Human 3.6M, 300-VW, and SURREAL datasets.

Handles missing data entries effectively.

Uses a simple network architecture with superior results.

Abstract

We propose a novel framework for training neural networks which is capable of learning 3D information of non-rigid objects when only 2D annotations are available as ground truths. Recently, there have been some approaches that incorporate the problem setting of non-rigid structure-from-motion (NRSfM) into deep learning to learn 3D structure reconstruction. The most important difficulty of NRSfM is to estimate both the rotation and deformation at the same time, and previous works handle this by regressing both of them. In this paper, we resolve this difficulty by proposing a loss function wherein the suitable rotation is automatically determined. Trained with the cost function consisting of the reprojection error and the low-rank term of aligned shapes, the network learns the 3D structures of such objects as human skeletons and faces during the training, whereas the testing is done in a single-frame basis. The proposed method can handle inputs with missing entries and experimental results validate that the proposed framework shows superior reconstruction performance to the state-of-the-art method on the Human 3.6M, 300-VW, and SURREAL datasets, even though the underlying network structure is very simple.