Building a representation of newly encountered faces: A role for context?
Bibliographic record
Abstract
Recognizing identity in naturalistic images of unfamiliar faces is challenging (Jenkins et al., 2011); two images of the same person often are misperceived as belonging to different people and images of different people often are misperceived as belonging to the same person. Thus, face learning involves increased tolerance of within-person variability in appearance and improved discrimination. We examined the process by which a perceiver determines the range of inputs that are attributable to a newly learned identity, such that novel images of that identity are recognized while those of a similar identity are excluded. Based on Tanaka's (2007) hypothesis that each identity is represented by an attractor field in multi-dimensional face space, the size of which is constrained by nearest neighbors, we predicted that learning a new face in the context of a similar identity would facilitate learning. In Experiment 1, participants (n=40) sorted 45 ambient images of three identities (15/identity) in the learning phase; two identities were similar (near neighbours) and one was dissimilar (far neighbour). In the test phase, participants identified new images of the learned identities when intermixed with novel identities. Performance (d', hits, FAs) did not vary for near vs. far neighbours, ps>0.15—perhaps because accuracy approached ceiling. In Experiment 2, participants (to date, n=24) sorted only 15 images to capture the representation of identity earlier in the learning process. Performance was worse in Experiment 2, p< 0.001; participants made fewer hits and more false alarms. Nonetheless, performance did not vary for near vs. far neighbours, ps>0.21. Collectively our findings confirm that identity learning involves both improved recognition of new instances (increased tolerance of variability) and improved discrimination. To date, though, we find no evidence that this learning is best accounted for by Valentine's (1991) multi-dimensional face space model, calling for revised theories of face recognition. Meeting abstract presented at VSS 2018
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.006 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".