notesum.ai

Published at November 6

Aligning Characteristic Descriptors with Images for Human-Expert-like Explainability

cs.CV

cs.AI

Released Date: November 6, 2024

Authors: Bharat Chandra Yalavarthi¹, Nalini Ratha¹

Aff.: ¹University at Buffalo

Arxiv: http://arxiv.org/abs/2411.04008v1

Refer to caption

Dataset	Proposed (Explainable)	Baseline (Black box - SOTA)
LFW [14]	98.38%	99.82%
CFP-FP [1]	90.88%	98.49%
CALFW [43]	89.36%	96.08%
AgeDB [29]	81.18%	98.05%
CPLFW [42]	87.71%	93.13%
Average	89.50%	97.11%