Redemption Score: A Multi-Modal Evaluation Framework for Image Captioning via Distributional, Perceptual, and Linguistic Signal Triangulation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dahal, Ashim, Ghimire, Ankit, Murad, Saydul Akbar, Rahimi, Nick
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918146829451264
author Dahal, Ashim
Ghimire, Ankit
Murad, Saydul Akbar
Rahimi, Nick
author_facet Dahal, Ashim
Ghimire, Ankit
Murad, Saydul Akbar
Rahimi, Nick
contents Evaluating image captions requires cohesive assessment of both visual semantics and language pragmatics, which is often not entirely captured by most metrics. We introduce Redemption Score(RS), a novel hybrid framework that ranks image captions by triangulating three complementary signals: (1) Mutual Information Divergence (MID) for global image-text distributional alignment, (2) DINO-based perceptual similarity of cycle-generated images for visual grounding, and (3) LLM Text Embeddings for contextual text similarity against human references. A calibrated fusion of these signals allows RS to offer a more holistic assessment. On the Flickr8k benchmark, RS achieves a Kendall-$τ$ of 58.42, outperforming most prior methods and demonstrating superior correlation with human judgments without requiring task-specific training. Our framework provides a more robust and nuanced evaluation by thoroughly examining both the visual accuracy and text quality together, with consistent performance across Conceptual Captions and MS COCO.
format Preprint
id arxiv_https___arxiv_org_abs_2505_16180
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Redemption Score: A Multi-Modal Evaluation Framework for Image Captioning via Distributional, Perceptual, and Linguistic Signal Triangulation
Dahal, Ashim
Ghimire, Ankit
Murad, Saydul Akbar
Rahimi, Nick
Computer Vision and Pattern Recognition
Computation and Language
Evaluating image captions requires cohesive assessment of both visual semantics and language pragmatics, which is often not entirely captured by most metrics. We introduce Redemption Score(RS), a novel hybrid framework that ranks image captions by triangulating three complementary signals: (1) Mutual Information Divergence (MID) for global image-text distributional alignment, (2) DINO-based perceptual similarity of cycle-generated images for visual grounding, and (3) LLM Text Embeddings for contextual text similarity against human references. A calibrated fusion of these signals allows RS to offer a more holistic assessment. On the Flickr8k benchmark, RS achieves a Kendall-$τ$ of 58.42, outperforming most prior methods and demonstrating superior correlation with human judgments without requiring task-specific training. Our framework provides a more robust and nuanced evaluation by thoroughly examining both the visual accuracy and text quality together, with consistent performance across Conceptual Captions and MS COCO.
title Redemption Score: A Multi-Modal Evaluation Framework for Image Captioning via Distributional, Perceptual, and Linguistic Signal Triangulation
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2505.16180