Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB

Medical Infant Motion Analysis


Infant motion analysis enables early detection of neurodevelopmental disorders like cerebral palsy (CP). The quality of spontaneous movements, in particular of the general movements (GMs), at the corrected age of 2-4 months accurately reflects the state of the infant's nervous system. The general movement assessment (GMA) method achieves the highest reliability for the detection of CP at an early age. In order to remove the human variability and the effort of regular training of GMA experts, we aim at automating medical infant motion analysis.

The Skinned Multi-Infant Linear model (SMIL)

Learning an Infant Body Model from RGB-D Data for Accurate Full Body Motion Analysis
Hesse, N., Pujades, S., Romero, J., Black, M. J., Bodensteiner, C., Arens, M., Hofmann, U. G., Tacke, U., Hadders-Algra, M., Weinberger, R., Müller-Felber, W., Schroeder, A. S.
In International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), September 2018
pdf supplementary video DOI bib

The SMIL model is available for research purposes.
To get access to the SMIL model, please fill the following form or send us an email with your name and affiliation.




The Moving INfants In RGB-D (MINI-RGBD) data set

Computer Vision for Medical Infant Motion Analysis: State of the Art and RGB-D Data Set
Hesse, N., Bodensteiner, C., Arens, M., Hofmann, U. G., Weinberger, R., Schroeder, A. S.
In European Conference on Computer Vision Workshops (ECCVW), September 2018
pdf bib video

By downloading and/or using the data set, you agree to the license terms, which can be found here.
Note: The infants shown in the images above / the video / the paper / the data set were created using the SMIL model with generated textures and shapes and therefore do not depict any existing infants.

To request the data set, please fill the following form.





Nikolas Hesse

Further Publications


Body Pose Estimation in Depth Images for Infant Motion Analysis
Hesse, N., Schröder, A. S., Müller-Felber, W., Bodensteiner, C., Arens, M., Hofmann, U. G.
39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 2017
pdf DOI bib

Markerless Motion Analysis for Early Detection of Infantile Movement Disorders
Hesse, N., Schroeder, A. S., Müller-Felber, W., Bodensteiner, C., Arens, M., Hofmann, U. G.
EMBEC & NBC 2017: Joint Conference of the European Medical and Biological Engineering Conference (EMBEC) and the Nordic-Baltic Conference on Biomedical Engineering and Medical Physics (NBC) 2017, Springer Singapore
pdf DOI bib

Entwicklungsneurologie - vernetzte Medizin und neue Perspektiven
Tacke, U., Weigand-Brunnhölzl, H., Hilgendorff, A., Giese, R. M., Flemmer, A. W., König, H., Warken-Madelung, B., Arens, M., Hesse, N., Schroeder, A. S.
Der Nervenarzt, 2017
DOI bib


Estimating Body Pose of Infants in Depth Images Using Random Ferns
Hesse, N., Stachowiak, G., Breuer, T., Arens, M.
IEEE International Conference on Computer Vision Workshop (ICCVW) 2015
pdf DOI bib