Unsupervised Deep Shape from Template

This paper presents Unsupervised Deep Shape from Template (UDSfT), a novel method that leverages deep neural networks (DNNs) for reconstructing the 3D surface of an object using a single image. More specifically, the reconstruction of isometric deformable objects is achieved in the proposed UDSfT method via a DNN-based template-based framework. Unlike previous approaches that leverage supervised learning, the proposed UDSfT method leverages the notion of unsupervised learning to overcome this obstacle and provide real-time 3D reconstruction. More specifically, UDSfT achieves this via an unsupervised structure that leverages a combination of real-data and synthetic data. Experimental results show that the proposed UDSfT method outperforms the state-of-the-art Shape from Template methods in object 3D reconstruction.

[1]  Didier Stricker,et al.  HDM-Net: Monocular Non-Rigid 3D Reconstruction with Learned Deformation Model , 2018, EuroVR.

[2]  Pascal Fua,et al.  Monocular 3D Reconstruction of Locally Textured Surfaces , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Daniel Pizarro-Perez,et al.  Shape-from-Template , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Richard Szeliski,et al.  Computer Vision - Algorithms and Applications , 2011, Texts in Computer Science.

[5]  Jin Liu,et al.  3D Feature Constrained Reconstruction for Low-Dose CT Imaging , 2018, IEEE Transactions on Circuits and Systems for Video Technology.

[6]  Rob Fergus,et al.  Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-scale Convolutional Architecture , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[7]  Zohreh Azimifar,et al.  Fast shape-from-template using local features , 2017, Machine Vision and Applications.

[8]  Adrien Bartoli,et al.  Inextensible Non-Rigid Structure-from-Motion by Second-Order Cone Programming , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9]  Pascal Fua,et al.  Template-Based Monocular 3D Shape Recovery Using Laplacian Meshes , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Pascal Fua,et al.  Laplacian Meshes for Monocular 3D Shape Recovery , 2012, ECCV.

[11]  Xiaogang Wang,et al.  T-CNN: Tubelets With Convolutional Neural Networks for Object Detection From Videos , 2016, IEEE Transactions on Circuits and Systems for Video Technology.

[12]  Xin Pan,et al.  A hybrid MLP-CNN classifier for very fine resolution remotely sensed image classification , 2017, ISPRS Journal of Photogrammetry and Remote Sensing.

[13]  Debdeep Banerjee,et al.  Robotic Arm Based 3D Reconstruction Test Automation , 2018, IEEE Access.