MétaCan
Menu
Back to cohort
Record W2751750293 · doi:10.1167/17.10.509

Extensive training of orientation filtered textures increases generalization of learning

2017· article· en· W2751750293 on OpenAlexaff
Ali Hashemi, Allison B. Sekuler, Patrick Bennett

Bibliographic record

VenueJournal of Vision · 2017
Typearticle
Languageen
FieldPsychology
TopicLanguage, Metaphor, and Cognition
Canadian institutionsMcMaster University
Fundersnot available
KeywordsContext (archaeology)GeneralizationArtificial intelligenceOrientation (vector space)Texture (cosmology)Computer scienceIdentification (biology)Task (project management)PerceptionPattern recognition (psychology)Context effectMachine learningComputer visionPsychologyMathematicsGeometryImage (mathematics)NeuroscienceGeographyBiology

Abstract

fetched live from OpenAlex

Previously, we (Hashemi et al., VSS 2015 & VSS 2016) investigated perceptual learning in a texture identification task using textures that contain diagnostic information (Target) in one orientation band and non-diagnostic information (Context) in a perpendicular orientation band. We found that, compared to Target-alone (i.e., no Context) textures, training in a 1-of-6 identification task with Target+Context textures resulted in lower accuracy, less learning, but greater generalization of learning. Specifically, training with Target+Context patterns generalized to familiar Target-alone textures, but training with Target-alone stimuli did not generalize to familiar Target+Context textures. Nevertheless, even with Target+Context training, we found no evidence of generalization to novel targets, regardless of context. Here we investigated whether greater generalization of learning could be obtained by significantly increasing the amount of training with Target+Context stimuli from 960 to 4200 trials. Before and after training, we assessed identification accuracy on trained and novel Targets with and without Context. We also tested accuracy on textures where the Target and Context orientations were swapped, using both novel and trained Targets. Results varied across observers: During training, accuracy increased by at least 50% in half of the participants, but only by ~20% in the others. Our post-training assessment found 1) all participants improved on the trained Target+Context textures; and some participants generalized learning to 2) familiar and novel Target-alone textures; 3) novel Target+Context textures; and/or 4) orientation-swapped Target+Context textures. Finally, the different patterns of generalization were not related in any simple way to the change in accuracy that occurred during training. Our results indicate that perceptual learning of orientation filtered textures varies significantly across individuals, but that it can be generalized to novel and familiar targets in novel contexts. These findings may have implications for perceptual learning in applied settings in which generalization of learning is a critical component of training. Meeting abstract presented at VSS 2017

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 imitation

Not 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.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.002
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: Bench or experimental
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.005
Threshold uncertainty score0.017

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0000.002
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0000.001
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0050.001

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.

Opus teacher head0.038
GPT teacher head0.363
Teacher spread0.325 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

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".

Quick stats

Citations0
Published2017
Admission routes1
Has abstractyes

Explore more

Same venueJournal of VisionSame topicLanguage, Metaphor, and CognitionFrench-language works237,207