SPEC: Seeing People in the Wild with an Estimated Camera

TL;DR

SPEC introduces a novel method for in-the-wild 3D human pose and shape estimation by accurately estimating camera parameters from a single image, significantly improving reconstruction accuracy over previous approaches that assumed fixed camera settings.

Contribution

The paper presents the first in-the-wild 3D human pose and shape estimation method that estimates camera parameters directly from a single image, enhancing reconstruction accuracy.

Findings

Outperforms prior methods on 3DPW benchmark

Creates new synthetic and in-the-wild datasets with camera calibration

Improves 3D human body reconstruction accuracy

Abstract

Due to the lack of camera parameter information for in-the-wild images, existing 3D human pose and shape (HPS) estimation methods make several simplifying assumptions: weak-perspective projection, large constant focal length, and zero camera rotation. These assumptions often do not hold and we show, quantitatively and qualitatively, that they cause errors in the reconstructed 3D shape and pose. To address this, we introduce SPEC, the first in-the-wild 3D HPS method that estimates the perspective camera from a single image and employs this to reconstruct 3D human bodies more accurately. First, we train a neural network to estimate the field of view, camera pitch, and roll given an input image. We employ novel losses that improve the calibration accuracy over previous work. We then train a novel network that concatenates the camera calibration to the image features and uses these together to regress 3D body shape and pose. SPEC is more accurate than the prior art on the standard benchmark (3DPW) as well as two new datasets with more challenging camera views and varying focal lengths. Specifically, we create a new photorealistic synthetic dataset (SPEC-SYN) with ground truth 3D bodies and a novel in-the-wild dataset (SPEC-MTP) with calibration and high-quality reference bodies. Both qualitative and quantitative analysis confirm that knowing camera parameters during inference regresses better human bodies. Code and datasets are available for research purposes at https://spec.is.tue.mpg.de.