Good Scores, Bad Data: A Metric for Multimodal Coherence

• URL: http://arxiv.org/abs/2603.25924v1
• Date: Thu, 26 Mar 2026 21:30:34 GMT
• Title: Good Scores, Bad Data: A Metric for Multimodal Coherence
• Abstract summary: We introduce the Multimodal Coherence Score (MCS), a metric that evaluates fusion quality independent of any downstream model.<n>We evaluate on 1,000 Visual Genome images using DETR, CLIP, and ViLT, and validate on 150 COCO images with no retraining.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Multimodal AI systems are evaluated by downstream task accuracy, but high accuracy does not mean the underlying data is coherent. A model can score well on Visual Question Answering (VQA) while its inputs contradict each other. We introduce the Multimodal Coherence Score (MCS), a metric that evaluates fusion quality independent of any downstream model. MCS decomposes coherence into four dimensions, identity, spatial, semantic, and decision, with weights learned via Nelder-Mead optimization. We evaluate on 1,000 Visual Genome images using DETR, CLIP, and ViLT, and validate on 150 COCO images with no retraining. Across three fusion architectures, MCS discriminates quality with higher sensitivity than task accuracy alone (Spearman rho = 0.093 vs. 0.071). Perturbation experiments confirm each dimension responds independently to its failure mode with zero cross-talk. MCS is lightweight, requires no human annotation, and tells you not just that something broke, but what broke.

Related papers

• qZACH-ViT: Quantization-Aware Intrinsic Explanations with Recursive Attribution-Stabilized Optimization [0.0]
  We introduce qZACH-ViT, a quantization-aware extension of the zero-token-free ZACH-ViT backbone.<n>We also introduce Recursive Attribution-Stabilized Optimization (RASO), which norm-matches classification and attribution gradients.<n>We evaluate four controlled conditions on seven MedMNIST datasets using 50 training images per class and ten fixed seeds.
  arXiv  Detail & Related papers  (2026-07-16T19:51:48Z)
• VendorBench-100: A Unified Cross-Paradigm Benchmark for Deepfake Image Detection [0.04736448323490553]
  VendorBench-100 is a cross-paradigm benchmark for deepfake image detection.<n>It evaluates 36 representative models using a single adversarial 100-image corpus, a unified output schema, and a common evaluation framework.<n>Our evaluation shows that commercial APIs achieve the strongest median performance, followed by vision LLMs and open-source detectors.
  arXiv  Detail & Related papers  (2026-07-07T13:22:00Z)
• Average Rankings Mask Per-Subject Optimality: A Friedman-Nemenyi Benchmark of EEG Motor-Imagery BCI Decoders [0.0]
  We evaluate 1,056 decoding configurations and 340,000 subject-level models.<n>Every model is fit and tested within a single session of a single participant.<n>Even in the easiest regime, no single pipeline dominates.
  arXiv  Detail & Related papers  (2026-06-23T10:26:27Z)
• EnactToM: An Evolving Benchmark for Functional Theory of Mind in Embodied Agents [75.01735520608075]
  Existing benchmarks mostly test literal Theory of Mind (ToM) by asking direct belief questions.<n>We introduce EnactToM, an evolving benchmark of 300 embodied multi-agent tasks set in a 3D household.<n>On the hard split, all seven evaluated frontier models score 0.0% Pass3 on functional task completion, while averaging 45.0% on literal belief probes.
  arXiv  Detail & Related papers  (2026-05-11T00:04:19Z)
• Metric Unreliability in Multimodal Machine Unlearning: A Systematic Analysis and Principled Unified Score [14.579552536669217]
  Unlearning in Vision-Language Models (VLMs) is required for compliance with the General Data Protection Regulation Code.<n>We present the first study of reliability in multimodal unlearning.<n>We introduce the Unified Quality Score (UQS), a metric with weights derived from each metric's Spearman correlation with the metric distance.
  arXiv  Detail & Related papers  (2026-05-04T04:13:00Z)
• From Global to Granular: Revealing IQA Model Performance via Correlation Surface [83.65597122328133]
  We present textbfGranularity-Modulated Correlation (GMC), which provides a structured, fine-grained analysis of IQA performance.<n>GMC includes a textbfDistribution Regulator that regularizes correlations to mitigate biases from non-uniform quality distributions.<n>Experiments on standard benchmarks show that GMC reveals performance characteristics invisible to scalar metrics, offering a more informative and reliable paradigm for analyzing, comparing, and deploying IQA models.
  arXiv  Detail & Related papers  (2026-01-29T13:55:26Z)
• Consistency-Aware Padding for Incomplete Multi-Modal Alignment Clustering Based on Self-Repellent Greedy Anchor Search [8.500604038628648]
  Multimodal representation is faithful and highly effective in describing real-world data samples' characteristics.<n> Existing research has not effectively addressed the issue of filling missing data in scenarios where multiview data are both imbalanced and misaligned.<n>We propose the Consistency-Aware Padding for Incomplete Multimodal Alignment Clustering Based on Self-Repellent Greedy Anchor Search.
  arXiv  Detail & Related papers  (2025-07-05T06:43:35Z)
• SPARC: Score Prompting and Adaptive Fusion for Zero-Shot Multi-Label Recognition in Vision-Language Models [74.40683913645731]
  Zero-shot multi-label recognition (MLR) with Vision-Language Models (VLMs) faces significant challenges without training data, model tuning, or architectural modifications.<n>Our work proposes a novel solution treating VLMs as black boxes, leveraging scores without training data or ground truth.<n>Analysis of these prompt scores reveals VLM biases and AND''/OR' signal ambiguities, notably that maximum scores are surprisingly suboptimal compared to second-highest scores.
  arXiv  Detail & Related papers  (2025-02-24T07:15:05Z)
• MedSegMamba: 3D CNN-Mamba Hybrid Architecture for Brain Segmentation [15.514511820130474]
  We develop a 3D patch-based hybrid CNN-Mamba model for subcortical brain segmentation. Our model's performance was validated against several benchmarks.
  arXiv  Detail & Related papers  (2024-09-12T02:19:19Z)
• Adaptive Feature Selection for No-Reference Image Quality Assessment by Mitigating Semantic Noise Sensitivity [55.399230250413986]
  We propose a Quality-Aware Feature Matching IQA Metric (QFM-IQM) to remove harmful semantic noise features from the upstream task.<n>Our approach achieves superior performance to the state-of-the-art NR-IQA methods on eight standard IQA datasets.
  arXiv  Detail & Related papers  (2023-12-11T06:50:27Z)
• VERITE: A Robust Benchmark for Multimodal Misinformation Detection Accounting for Unimodal Bias [17.107961913114778]
  multimodal misinformation is a growing problem on social media platforms. In this study, we investigate and identify the presence of unimodal bias in widely-used MMD benchmarks. We introduce a new method -- termed Crossmodal HArd Synthetic MisAlignment (CHASMA) -- for generating realistic synthetic training data.
  arXiv  Detail & Related papers  (2023-04-27T12:28:29Z)
• Few-Shot Non-Parametric Learning with Deep Latent Variable Model [50.746273235463754]
  We propose Non-Parametric learning by Compression with Latent Variables (NPC-LV) NPC-LV is a learning framework for any dataset with abundant unlabeled data but very few labeled ones. We show that NPC-LV outperforms supervised methods on all three datasets on image classification in low data regime.
  arXiv  Detail & Related papers  (2022-06-23T09:35:03Z)
• Twitter-COMMs: Detecting Climate, COVID, and Military Multimodal Misinformation [83.2079454464572]
  This paper describes our approach to the Image-Text Inconsistency Detection challenge of the DARPA Semantic Forensics (SemaFor) Program. We collect Twitter-COMMs, a large-scale multimodal dataset with 884k tweets relevant to the topics of Climate Change, COVID-19, and Military Vehicles. We train our approach, based on the state-of-the-art CLIP model, leveraging automatically generated random and hard negatives.
  arXiv  Detail & Related papers  (2021-12-16T03:37:20Z)
• Comprehensive Graph-conditional Similarity Preserving Network for Unsupervised Cross-modal Hashing [97.44152794234405]
  Unsupervised cross-modal hashing (UCMH) has become a hot topic recently. In this paper, we devise a deep graph-neighbor coherence preserving network (DGCPN) DGCPN regulates comprehensive similarity preserving losses by exploiting three types of data similarities.
  arXiv  Detail & Related papers  (2020-12-25T07:40:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.