CrowdRec: 3D Crowd Reconstruction from Single Color Images
Buzhen Huang, Jingyi Ju, Yangang Wang

TL;DR
CrowdRec introduces a novel crowd-constrained optimization approach that enhances 3D crowd reconstruction from single color images by refining single-person mesh recovery with crowd features, addressing occlusion and spatial distribution challenges.
Contribution
The paper proposes a crowd-constrained optimization method that improves 3D crowd reconstruction accuracy from monocular images by integrating crowd features into single-person mesh recovery.
Findings
Achieves accurate 3D body poses and shapes in crowded scenes
Effectively refines single-person mesh recovery using crowd constraints
Addresses occlusion and spatial distribution issues in 3D crowd reconstruction
Abstract
This is a technical report for the GigaCrowd challenge. Reconstructing 3D crowds from monocular images is a challenging problem due to mutual occlusions, server depth ambiguity, and complex spatial distribution. Since no large-scale 3D crowd dataset can be used to train a robust model, the current multi-person mesh recovery methods can hardly achieve satisfactory performance in crowded scenes. In this paper, we exploit the crowd features and propose a crowd-constrained optimization to improve the common single-person method on crowd images. To avoid scale variations, we first detect human bounding-boxes and 2D poses from the original images with off-the-shelf detectors. Then, we train a single-person mesh recovery network using existing in-the-wild image datasets. To promote a more reasonable spatial distribution, we further propose a crowd constraint to refine the single-person network…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsVideo Surveillance and Tracking Methods · Human Pose and Action Recognition · Advanced Image Processing Techniques
