Bridging the Missing-Modality Gap: Improving Text-Only Calibration of Vision Language Models

• URL: http://arxiv.org/abs/2605.12517v1
• Date: Fri, 03 Apr 2026 10:03:02 GMT
• Title: Bridging the Missing-Modality Gap: Improving Text-Only Calibration of Vision Language Models
• Abstract summary: Vision-language models (VLMs) are often deployed on text-only inputs, although they are trained with images.<n>We find that removing the vision modality causes large drops in accuracy and severe miscalibration, and the model does not behave like its original language backbone under text-only prompting.<n>We propose the Latent Imagination Module (LIM), a lightweight cross-attention module that predicts imagined latent embeddings from textual input and feeds them into a frozen VLM backbone without pixel-level image synthesis.
• Score: 14.752088383510788
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-language models (VLMs) are often deployed on text-only inputs, although they are trained with images. We find that removing the vision modality causes large drops in accuracy and severe miscalibration, and the model does not behave like its original language backbone under text-only prompting. This failure is not explained only by missing semantic information. Even when text descriptions preserve key content, confidence becomes unreliable, while adding a visual signal through generated images partially restores accuracy and calibration. We propose the Latent Imagination Module (LIM), a lightweight cross-attention module that predicts imagined latent embeddings from textual input and feeds them into a frozen VLM backbone without pixel-level image synthesis. Across text-only benchmarks, unseen tasks, and missing-image scenarios, LIM improves accuracy and reduces calibration error. These results suggest that latent modality completion is a practical approach for reliable VLM inference under missing-modality.

Related papers

• T-IMPACT: A Severity-Aware Benchmark for Contextual Image-Text Manipulation [9.049034101566642]
  We introduce T-context, a first-release severity-aware benchmark for manipulated news-style image-text pairs.<n>T- contains 98,786 examples spanning pristine, image-only, text-only, and joint manipulations.<n>The pipeline extracts semantic anchors, grounds them spatially, performs localized image edits and constrained caption rewrites.
  arXiv  Detail & Related papers  (2026-06-21T05:07:20Z)
• HNC: Leveraging Hard Negative Captions towards Models with Fine-Grained Visual-Linguistic Comprehension Capabilities [11.816108209604211]
  Hard Negative Captions (HNC) is an automatically created dataset containing foiled hard negative captions for ITM training.<n>We show the effectiveness of training on HNC by improving the models' zero-shot capabilities in detecting mismatches on diagnostic tasks and performing robustly under noisy visual input scenarios.
  arXiv  Detail & Related papers  (2026-05-06T14:01:47Z)
• Evaluating Remote Sensing Image Captions Beyond Metric Biases [20.77920438007019]
  We introduce ReconScore, a novel reference-free evaluation metric.<n>We use ReconScore to assess caption quality by its capability to reconstruct the original visual elements solely from the generated text.<n>By employing ReconScore as a self-correction mechanism, we iteratively refine the semantic precision of MLLM outputs without any computational fine-tuning overhead.
  arXiv  Detail & Related papers  (2026-04-22T12:28:04Z)
• Reading, Not Thinking: Understanding and Bridging the Modality Gap When Text Becomes Pixels in Multimodal LLMs [23.958966900531692]
  Multimodal large language models (MLLMs) can process text presented as images, yet they often perform worse than when the same content is provided as textual tokens.<n>We systematically diagnose this "modality gap" by evaluating seven MLLMs across seven benchmarks in five input modes.
  arXiv  Detail & Related papers  (2026-03-10T02:14:23Z)
• FineGRAIN: Evaluating Failure Modes of Text-to-Image Models with Vision Language Model Judges [85.24983823102262]
  We propose a structured methodology for evaluating text-to-image (T2I) models and vision language models (VLMs)<n>We test whether VLMs can identify 27 specific failure modes in the images generated by T2I models conditioned on challenging prompts.<n>Our findings suggest that current metrics are insufficient to capture these nuanced errors.
  arXiv  Detail & Related papers  (2025-12-01T19:46:03Z)
• PoSh: Using Scene Graphs To Guide LLMs-as-a-Judge For Detailed Image Descriptions [55.95282725491425]
  PoSh is a metric for detailed image description that uses scene graphs as structured rubrics to guide LLMs-as-a-Judge.<n>PoSh is replicable, interpretable and a better proxy for human raters than existing metrics.<n>We show that PoSh achieves stronger correlations with the human judgments in DOCENT than the best open-weight alternatives.
  arXiv  Detail & Related papers  (2025-10-21T20:30:20Z)
• Extract Free Dense Misalignment from CLIP [7.0247398611254175]
  This work proposes a novel approach, dubbed CLIP4DM, for detecting dense misalignments from pre-trained CLIP.<n>We revamp the gradient-based attribution computation method, enabling negative gradient of individual text tokens to indicate misalignment.<n>Our method demonstrates state-of-the-art performance among zero-shot models and competitive performance with fine-tuned models.
  arXiv  Detail & Related papers  (2024-12-24T12:51:05Z)
• HoVLE: Unleashing the Power of Monolithic Vision-Language Models with Holistic Vision-Language Embedding [91.0552157725366]
  This paper presents a novel high-performance monolithic VLM named HoVLE.<n>It converts visual and textual inputs into a shared space, allowing LLMs to process images in the same way as texts.<n>Our experiments show that HoVLE achieves performance close to leading compositional models on various benchmarks.
  arXiv  Detail & Related papers  (2024-12-20T18:59:59Z)
• Looking Beyond Text: Reducing Language bias in Large Vision-Language Models via Multimodal Dual-Attention and Soft-Image Guidance [67.26434607115392]
  Large vision-language models (LVLMs) have achieved impressive results in various vision-language tasks. LVLMs suffer from hallucinations caused by language bias, leading to diminished focus on images and ineffective visual comprehension. We propose LACING to address the language bias of LVLMs with muLtimodal duAl-attention meChanIsm (MDA) aNd soft-image Guidance (IFG)
  arXiv  Detail & Related papers  (2024-11-21T16:33:30Z)
• Debiasing Vison-Language Models with Text-Only Training [15.069736314663352]
  We propose a Text-Only Debiasing framework called TOD, leveraging a text-as-image training paradigm to mitigate visual biases. To address the limitations, we propose a Text-Only Debiasing framework called TOD, leveraging a text-as-image training paradigm to mitigate visual biases.
  arXiv  Detail & Related papers  (2024-10-12T04:34:46Z)
• FINEMATCH: Aspect-based Fine-grained Image and Text Mismatch Detection and Correction [66.98008357232428]
  We propose FineMatch, a new aspect-based fine-grained text and image matching benchmark. FineMatch focuses on text and image mismatch detection and correction. We show that models trained on FineMatch demonstrate enhanced proficiency in detecting fine-grained text and image mismatches.
  arXiv  Detail & Related papers  (2024-04-23T03:42:14Z)
• Do LVLMs Understand Charts? Analyzing and Correcting Factual Errors in Chart Captioning [90.13978453378768]
  We introduce a comprehensive typology of factual errors in generated chart captions. A large-scale human annotation effort provides insight into the error patterns and frequencies in captions crafted by various chart captioning models. Our analysis reveals that even state-of-the-art models, including GPT-4V, frequently produce captions laced with factual inaccuracies.
  arXiv  Detail & Related papers  (2023-12-15T19:16:21Z)

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.