On standard face benchmarks, domain-specific face recognition models beat zero-shot foundation models, adding context or fusing scores improves performance at low false-match rates, and GPT-4o can explain and sometimes correct AdaFace's errors.
Technical report, National Institute of Standards and Technology, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Foundation versus Domain-specific Models: Performance Comparison, Fusion, and Explainability in Face Recognition
On standard face benchmarks, domain-specific face recognition models beat zero-shot foundation models, adding context or fusing scores improves performance at low false-match rates, and GPT-4o can explain and sometimes correct AdaFace's errors.