Back

Validity of Low-Dimensional Representation of Whole-Body Movements

Wakabayashi, M.; Katayama, A.; Asai, T.

2026-09-04 neuroscience
10.64898/2026.08.31.746368 bioRxiv
Show abstract

Research on impression evaluation from whole-body movements has traditionally focused on the relationships between emotional perception or aesthetic evaluation and movement patterns characterized by local kinematic features and inter-limb coordination. However, because observers perceive the entire body simultaneously when evaluating emotions or aesthetics in real-world viewing situations, impressions derived from whole-body movements should also be investigated. This requires a low-dimensional representation that summarizes whole-body information in an interpretable manner. In the present study, we evaluated whether the state space generated by multidimensional scaling (MDS) provides a valid representation for summarizing whole-body movement. Specifically, using exercises from Radio Calisthenics No. 1, we examined whether distances between points in the MDS state space primarily reflected exercise category rather than individual differences. The results showed that distances in the state space reflected exercise category more strongly than performer identity, suggesting that MDS captures similarities among postures of the same exercise. However, the patterns of distances between conditions varied across exercise categories. These findings suggest that MDS-based state-space representations provide a valid and interpretable approach for summarizing whole-body movements. Nevertheless, further validation using a wider variety of whole-body movements and a more diverse participant population is required.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.