Generating synthetic images of human skeletal motion for pose and kinematics estimation tasks
Jere Lavikainen, David Pagnon, Mimmi K Liukkonen, Rami K Korhonen, Mikael J Turunen, Lauri Stenroth, Mika E Mononen · 2025
Read Generating synthetic images of human skeletal motion for pose and kinematics estimation tasks by Jere Lavikainen, David Pagnon, Mimmi K Liukkonen, Rami K Korhonen, Mikael J Turunen, Lauri Stenroth, Mika E Mononen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12711891.txt.
Description
Jere Lavikainen and co-authors build software that manufactures fake motion footage for training pose tracking systems. Markerless pose estimation promises cheap, accurate kinematics measurement, but labeled training data are scarce and hand marked by non-experts. Their program renders images straight from musculoskeletal models, with each frame automatically tagged with the true three-dimensional joint positions.
From a handful of skeletal poses it can spin unlimited images, swapping camera angles, backgrounds, body shapes, and skin and clothing textures. Because the labels come from the biomechanical model itself, they can outrun the accuracy of real photo sets, and new landmark points can be defined by editing the model's virtual marker set. The authors offer it as data factory for movement science.
Catalog record
| CREATOR | Jere Lavikainen, David Pagnon, Mimmi K Liukkonen, Rami K Korhonen, Mikael J Turunen, Lauri Stenroth, Mika E Mononen |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |