"Înţelegi Româneşte?'' A Recipe for Romanian Vision-Language Models

• URL: http://arxiv.org/abs/2605.31401v2
• Date: Mon, 01 Jun 2026 04:58:07 GMT
• Title: "Înţelegi Româneşte?'' A Recipe for Romanian Vision-Language Models
• Abstract summary: We present a systematic study of building a language-specific Vision-Language Models (VLMs) for Romanian.<n>We translate established English VLM training and evaluation corpora into Romanian, applying machine translation to textual annotations and to in-image text.<n>Using this data, we train and ablate a series of VLMs to isolate the contribution of vision backbones of varying scale and pretraining.<n>We also curate HoraVQA, a culturally native evaluation set grounded in Romanian everyday scenes.
• Score: 7.120569645707792
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-Language Models (VLMs) largely follow the text-only LLM trajectory, excelling on English benchmarks but sharply degrading on low-resource languages, where neither large-scale image-text corpora nor culturally grounded evaluations exist. We present a systematic study of building a language-specific VLM for Romanian, covering the full pipeline from data construction to architectural choices. We translate established English VLM training and evaluation corpora into Romanian, applying machine translation to textual annotations and to in-image text, preserving visual grounding while adapting the textual content. Using this data, we train and ablate a series of VLMs to isolate the contribution of (i) vision backbones of varying scale and pretraining, (ii) language backbones from multilingual to Romanian-adapted LLMs, and (iii) OCR-style image-text data. We further curate HoraVQA, a culturally native evaluation set grounded in Romanian everyday scenes. Romanian-adapted VLMs consistently outperform their same-sized counterparts and, across all evaluated benchmarks, even surpass models from the next larger size category.

Related papers

• A Pilot Study on Curator-Guided Multilingual Art Description for Blind and Low-Vision Audiences with Small Vision-Language Models [50.00033296815133]
  Blind and low-vision (BLV) audiences remain underserved by visual art descriptions.<n>Pilot study investigates curator-guided multilingual art description with Qwen2.5-VL-3B-Instruct for German, Romanian, and Serbian.
  arXiv  Detail & Related papers  (2026-05-29T09:47:45Z)
• Multilingual Training and Evaluation Resources for Vision-Language Models [2.5812054740604453]
  We introduce a new comprehensive suite of resources for Vision Language Models (VLMs) training and evaluation spanning five European languages.<n>We use a regeneration-translation paradigm that produces high-quality cross-lingual resources by combining curated synthetic generation and manual annotation.<n>Experiments show that using multilingual, multimodal examples for training VLMs aids is consistently beneficial on non-English benchmarks.
  arXiv  Detail & Related papers  (2026-04-20T14:42:47Z)
• Improving Romanian LLM Pretraining Data using Diversity and Quality Filtering [1.7705784090599055]
  Large Language Models (LLMs) have recently exploded in popularity, often matching or outperforming human abilities on many tasks.<n>One of the key factors in training LLMs is the availability and curation of high-quality data.<n>Data quality is especially crucial for under-represented languages, where high-quality corpora are scarce.
  arXiv  Detail & Related papers  (2025-11-02T21:41:49Z)
• TowerVision: Understanding and Improving Multilinguality in Vision-Language Models [56.775118098058506]
  TowerVision is a family of open multilingual vision-language models for both image-text and video-text tasks.<n>By incorporating visual and cultural context during fine-tuning, our models surpass existing approaches.<n>To support further research, we publicly release all models, data, and training recipes.
  arXiv  Detail & Related papers  (2025-10-22T17:02:48Z)
• A Culturally-Rich Romanian NLP Dataset from "Who Wants to Be a Millionaire?" Videos [44.99833362998488]
  Large Language Models (LLMs) demonstrate varying performance across languages and cultural contexts.<n>This study introduces a novel, culturally-rich, multilingual dataset derived from video recordings of the Romanian game show "Who Wants to Be a Millionaire?"
  arXiv  Detail & Related papers  (2025-06-06T11:21:38Z)
• "Vorbeşti Româneşte?" A Recipe to Train Powerful Romanian LLMs with English Instructions [40.64721381920061]
  We are the first to collect and translate a large collection of texts, instructions, and benchmarks and train, evaluate, and release open-source LLMs tailored for Romanian. We argue for the usefulness and high performance of RoLLMs by obtaining state-of-the-art results across the board.
  arXiv  Detail & Related papers  (2024-06-26T11:39:51Z)
• RC3: Regularized Contrastive Cross-lingual Cross-modal Pre-training [84.23022072347821]
  We propose a regularized cross-lingual visio-textual contrastive learning objective that constrains the representation proximity of weakly-aligned visio-textual inputs. Experiments on 5 downstream multi-modal tasks across 6 languages demonstrate the effectiveness of our proposed method.
  arXiv  Detail & Related papers  (2023-05-13T14:41:05Z)
• IGLUE: A Benchmark for Transfer Learning across Modalities, Tasks, and Languages [87.5457337866383]
  We introduce the Image-Grounded Language Understanding Evaluation benchmark. IGLUE brings together visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. We find that translate-test transfer is superior to zero-shot transfer and that few-shot learning is hard to harness for many tasks.
  arXiv  Detail & Related papers  (2022-01-27T18:53:22Z)
• 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)
• Room-Across-Room: Multilingual Vision-and-Language Navigation with Dense Spatiotemporal Grounding [75.03682706791389]
  We introduce Room-Across-Room (RxR), a new Vision-and-Language Navigation (VLN) dataset. RxR is multilingual (English, Hindi, and Telugu) and larger (more paths and instructions) than other VLN datasets. It emphasizes the role of language in VLN by addressing known biases in paths and eliciting more references to visible entities.
  arXiv  Detail & Related papers  (2020-10-15T18:01:15Z)

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.