Skip to Main content Skip to Navigation
Book sections

Learning an Infant Body Model from RGB-D Data for Accurate Full Body Motion Analysis

Abstract : Infant motion analysis enables early detection of neurode-velopmental disorders like cerebral palsy (CP). Diagnosis, however, is challenging, requiring expert human judgement. An automated solution would be beneficial but requires the accurate capture of 3D full-body movements. To that end, we develop a non-intrusive, low-cost, lightweight acquisition system that captures the shape and motion of infants. Going beyond work on modeling adult body shape, we learn a 3D Skinned Multi-Infant Linear body model (SMIL †) from noisy, low-quality, and incomplete RGB-D data. We demonstrate the capture of shape and motion with 37 infants in a clinical environment. Quantitative experiments show that SMIL faithfully represents the data and properly factorizes the shape and pose of the infants. With a case study based on general movement assessment (GMA), we demonstrate that SMIL captures enough information to allow medical assessment. SMIL provides a new tool and a step towards a fully automatic system for GMA.
Complete list of metadata

Cited literature [21 references]  Display  Hide  Download


https://hal.inria.fr/hal-02162177
Contributor : Sergi Pujades <>
Submitted on : Friday, June 21, 2019 - 3:27:43 PM
Last modification on : Monday, November 9, 2020 - 5:32:02 PM

Files

hesseMiccai18.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02162177, version 1

Citation

Nikolas Hesse, Sergi Pujades, Javier Romero, Michael Black, Christoph Bodensteiner, et al.. Learning an Infant Body Model from RGB-D Data for Accurate Full Body Motion Analysis. In A. Frangi, J. Schnabel, C. Davatzikos, C. Alberola-López, & G. Fichtinger (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI, pp.792-800, 2018. ⟨hal-02162177⟩

Share

Metrics

Record views

310

Files downloads

1060