A Multimodal Recaptioning Framework to Account for Perceptual Diversity in Multilingual Vision-Language Modeling

• URL: http://arxiv.org/abs/2504.14359v1
• Date: Sat, 19 Apr 2025 17:23:12 GMT
• Title: A Multimodal Recaptioning Framework to Account for Perceptual Diversity in Multilingual Vision-Language Modeling
• Authors: Kyle Buettner, Jacob Emmerson, Adriana Kovashka,
• Abstract summary: Machine translation of captions has pushed multilingual capabilities in vision-language models (VLMs)<n>Data comes mainly from English speakers, indicating a perceptual bias and lack of model flexibility.<n>We propose an LLM-based, multimodal recaptioning strategy that alters the object descriptions of English captions before translation.
• Score: 25.43735315887918
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: There are many ways to describe, name, and group objects when captioning an image. Differences are evident when speakers come from diverse cultures due to the unique experiences that shape perception. Machine translation of captions has pushed multilingual capabilities in vision-language models (VLMs), but data comes mainly from English speakers, indicating a perceptual bias and lack of model flexibility. In this work, we address this challenge and outline a data-efficient framework to instill multilingual VLMs with greater understanding of perceptual diversity. We specifically propose an LLM-based, multimodal recaptioning strategy that alters the object descriptions of English captions before translation. The greatest benefits are demonstrated in a targeted multimodal mechanism guided by native speaker data. By adding produced rewrites as augmentations in training, we improve on German and Japanese text-image retrieval cases studies (up to +3.5 mean recall overall, +4.7 on non-native error cases). We further propose a mechanism to analyze the specific object description differences across datasets, and we offer insights into cross-dataset and cross-language generalization.

Related papers

• Breaking Language Barriers in Visual Language Models via Multilingual Textual Regularization [9.349707150988893]
  We propose a continuous multilingual integration strategy that injects text-only multilingual data during visual instruction tuning.<n>Our approach significantly improves linguistic fidelity across languages without degradation in visual performance.
  arXiv  Detail & Related papers  (2025-03-28T16:26:52Z)
• Quantifying the Gaps Between Translation and Native Perception in Training for Multimodal, Multilingual Retrieval [28.589035749529955]
  We empirically show performance gaps between training on captions that come from native German perception and captions that have been either machine-translated or human-translated from English into German. While we achieve mean recall improvements (+1.3), gaps still remain, indicating an open area of future work for the community.
  arXiv  Detail & Related papers  (2024-10-02T20:47:53Z)
• Multilingual Diversity Improves Vision-Language Representations [66.41030381363244]
  Pre-training on this dataset outperforms using English-only or English-dominated datasets on ImageNet. On a geographically diverse task like GeoDE, we also observe improvements across all regions, with the biggest gain coming from Africa.
  arXiv  Detail & Related papers  (2024-05-27T08:08:51Z)
• Multi-modal Instruction Tuned LLMs with Fine-grained Visual Perception [63.03288425612792]
  We propose bfAnyRef, a general MLLM model that can generate pixel-wise object perceptions and natural language descriptions from multi-modality references. Our model achieves state-of-the-art results across multiple benchmarks, including diverse modality referring segmentation and region-level referring expression generation.
  arXiv  Detail & Related papers  (2024-03-05T13:45:46Z)
• CLARA: Multilingual Contrastive Learning for Audio Representation Acquisition [5.520654376217889]
  CLARA minimizes reliance on labelled data, enhancing generalization across languages. Our approach adeptly captures emotional nuances in speech, overcoming subjective assessment issues. It adapts to low-resource languages, marking progress in multilingual speech representation learning.
  arXiv  Detail & Related papers  (2023-10-18T09:31:56Z)
• VLIS: Unimodal Language Models Guide Multimodal Language Generation [23.094728230459125]
  We introduce Visual-Language models as Importance Sampling weights (VLIS) It combines the visual conditioning capability of vision-language models with the language understanding of unimodal text-only language models without further training. VLIS improves vision-language models on diverse tasks, including commonsense understanding and complex text generation.
  arXiv  Detail & Related papers  (2023-10-15T07:58:52Z)
• Multimodal Modeling For Spoken Language Identification [57.94119986116947]
  Spoken language identification refers to the task of automatically predicting the spoken language in a given utterance. We propose MuSeLI, a Multimodal Spoken Language Identification method, which delves into the use of various metadata sources to enhance language identification.
  arXiv  Detail & Related papers  (2023-09-19T12:21:39Z)
• DiMBERT: Learning Vision-Language Grounded Representations with Disentangled Multimodal-Attention [101.99313208598569]
  Vision-and-language (V-L) tasks require the system to understand both vision content and natural language. We propose DiMBERT (short for Disentangled Multimodal-Attention BERT), which applies separated attention spaces for vision and language. We show that DiMBERT sets new state-of-the-art performance on three tasks.
  arXiv  Detail & Related papers  (2022-10-28T23:00:40Z)
• What's in a Caption? Dataset-Specific Linguistic Diversity and Its Effect on Visual Description Models and Metrics [14.624063829492764]
  We find that caption diversity is a major driving factor behind the generation of generic and uninformative captions. We show that state-of-the-art models even outperform held-out ground truth captions on modern metrics.
  arXiv  Detail & Related papers  (2022-05-12T17:55:08Z)
• VidLanKD: Improving Language Understanding via Video-Distilled Knowledge Transfer [76.3906723777229]
  We present VidLanKD, a video-language knowledge distillation method for improving language understanding. We train a multi-modal teacher model on a video-text dataset, and then transfer its knowledge to a student language model with a text dataset. In our experiments, VidLanKD achieves consistent improvements over text-only language models and vokenization models.
  arXiv  Detail & Related papers  (2021-07-06T15:41:32Z)
• UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pre-training [52.852163987208826]
  UC2 is the first machine translation-augmented framework for cross-lingual cross-modal representation learning. We propose two novel pre-training tasks, namely Masked Region-to-Token Modeling (MRTM) and Visual Translation Language Modeling (VTLM) Our proposed framework achieves new state-of-the-art on diverse non-English benchmarks while maintaining comparable performance to monolingual pre-trained models on English tasks.
  arXiv  Detail & Related papers  (2021-04-01T08:30:53Z)

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.