3D-LMNet: Latent embedding matching for accurate and diverse 3D point cloud reconstruction from a single image

Mandikal, P and Navaneet, KL and Agarwal, M and Venkatesh, BR (2018) 3D-LMNet: Latent embedding matching for accurate and diverse 3D point cloud reconstruction from a single image. In: 29th British Machine Vision Conference, 3-6 September 2018, Newcastle.

PDF BMVC_2018.pdf - Published Version Restricted to Registered users only Download (8MB)

Abstract

3D reconstruction from single view images is an ill-posed problem. Inferring the hidden regions from self-occluded images is both challenging and ambiguous. We propose a two-pronged approach to address these issues. To better incorporate the data prior and generate meaningful reconstructions, we propose 3D-LMNet, a latent embedding matching approach for 3D reconstruction. We first train a 3D point cloud auto-encoder and then learn a mapping from the 2D image to the corresponding learnt embedding. To tackle the issue of uncertainty in the reconstruction, we predict multiple reconstructions that are consistent with the input view. This is achieved by learning a probablistic latent space with a novel view-specific 'diversity loss'. Thorough quantitative and qualitative analysis is performed to highlight the significance of the proposed approach. We outperform state-of-the-art approaches on the task of single-view 3D reconstruction on both real and synthetic datasets while generating multiple plausible reconstructions, demonstrating the generalizability and utility of our approach.

Item Type:	Conference Paper
Publication:	British Machine Vision Conference 2018, BMVC 2018
Publisher:	BMVA Press
Additional Information:	The copyright for this article belongs to BMVA Press.
Keywords:	Computer vision; Embeddings; Repair, 3D point cloud; 3D reconstruction; Diversity loss; Ill posed problem; Quantitative and qualitative analysis; Single-view 3D reconstruction; State-of-the-art approach; Synthetic datasets, Image reconstruction
Department/Centre:	Division of Interdisciplinary Sciences > Computational and Data Sciences
Date Deposited:	02 Dec 2022 09:24
Last Modified:	02 Dec 2022 09:24
URI:	https://eprints.iisc.ac.in/id/eprint/78165

Actions (login required)

View Item