Person identification across actions from biological motion
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
A significant amount of past research has studied person identification from point light displays of walking humans, investigating parameters such as viewing angle and the differential contributions of structural and kinematic information. However, little is known about the ability of human observers to generalize identity across different activities. In this study we use a same/different paradigm to compare observers' ability to identify point light displays within and across activities. We drew from a database of 100 motion-captured humans, each of which encompassed both walking and running activities. Subjects were shown successive paired stimuli and had to indicate whether the stimuli represented the same or different person. In either case, the two displays were at slightly different viewpoints. Two independent factors were examined: stimulus pairing (walker/walker, runner/runner, walker/runner) and information content (structural only, kinematic only, full information). For all information contents for stimulus pairing of matching activities (walker/walker, runner/runner) subjects performed significantly better than chance (t(5)=2.71, p0.05). The main effect of Pairing was significant (F(2, 30)=35.7, p[[lt]]0.001), with the walker/runner pairing being the most difficult. Information was not a significant factor. However, there was a significant interaction between Pairing and Information (F(4, 30)=4.03, p[[lt]]0.01) that manifested in performance on the runner/runner task in particular being better for full information than for structural or kinematic only. Results are discussed in light of a principal components-based linear model that estimates a runner time series from a given walker time series by equating principal component coordinates.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it