ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Unsupervised Cross-Modal Alignment for Multi-person 3D Pose Estimation

Kundu, JN and Revanur, A and Waghmare, GV and Venkatesh, RM and Babu, RV (2020) Unsupervised Cross-Modal Alignment for Multi-person 3D Pose Estimation. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 23-28 August 2020, Glasgow, pp. 35-52.

ECCV 2020.pdf - Published Version

Download (4MB) | Preview
Official URL: https://doi.org/10.1007/978-3-030-58601-0_3


We present a deployment friendly, fast bottom-up framework for multi-person 3D human pose estimation. We adopt a novel neural representation of multi-person 3D pose which unifies the position of person instances with their corresponding 3D pose representation. This is realized by learning a generative pose embedding which not only ensures plausible 3D pose predictions, but also eliminates the usual keypoint grouping operation as employed in prior bottom-up approaches. Further, we propose a practical deployment paradigm where paired 2D or 3D pose annotations are unavailable. In the absence of any paired supervision, we leverage a frozen network, as a teacher model, which is trained on an auxiliary task of multi-person 2D pose estimation. We cast the learning as a cross-modal alignment problem and propose training objectives to realize a shared latent space between two diverse modalities. We aim to enhance the model’s ability to perform beyond the limiting teacher network by enriching the latent-to-3D pose mapping using artificially synthesized multi-person 3D scene samples. Our approach not only generalizes to in-the-wild images, but also yields a superior trade-off between speed and performance, compared to prior top-down approaches. Our approach also yields state-of-the-art multi-person 3D pose estimation performance among the bottom-up approaches under consistent supervision levels.

Item Type: Conference Proceedings
Publication: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher: Springer Science and Business Media Deutschland GmbH
Additional Information: The copyright for this article belongs to the Authors.
Keywords: Alignment; Computer vision; Economic and social effects, 3D human pose estimation; 3D pose estimation; Alignment Problems; Bottom up approach; Bottom-up frameworks; Grouping operations; Neural representations; Top down approaches, Modal analysis
Department/Centre: Division of Interdisciplinary Sciences > Computational and Data Sciences
Date Deposited: 23 Jan 2023 10:07
Last Modified: 23 Jan 2023 10:07
URI: https://eprints.iisc.ac.in/id/eprint/79263

Actions (login required)

View Item View Item