---
license: apache-2.0
task_categories:
- translation
- question-answering
- text-generation
language:
- bn
- en
- hi
- ur
pretty_name: BanglaVerse
size_categories:
- 10K<n<100K
tags:
- multimodal
- bengali-culture
- dialects
- multilingual
configs:
- config_name: default
  data_files:
  - split: pure_bn
    path: data/pure_bn-*
  - split: english
    path: data/english-*
  - split: hindi
    path: data/hindi-*
  - split: urdu
    path: data/urdu-*
  - split: barishal
    path: data/barishal-*
  - split: chittagong
    path: data/chittagong-*
  - split: noakhali
    path: data/noakhali-*
  - split: rangpur
    path: data/rangpur-*
  - split: sylhet
    path: data/sylhet-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_id
    dtype: string
  - name: domain
    dtype: string
  - name: caption
    dtype: string
  - name: original_question
    dtype: string
  - name: question
    dtype: string
  - name: original_options
    list: string
  - name: options
    list: string
  - name: original_answer
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: pure_bn
    num_bytes: 395518633.728
    num_examples: 1152
  - name: english
    num_bytes: 395225870.728
    num_examples: 1152
  - name: hindi
    num_bytes: 395546867.728
    num_examples: 1152
  - name: urdu
    num_bytes: 395352490.728
    num_examples: 1152
  - name: barishal
    num_bytes: 395866114.728
    num_examples: 1152
  - name: chittagong
    num_bytes: 395857606.728
    num_examples: 1152
  - name: noakhali
    num_bytes: 395864142.728
    num_examples: 1152
  - name: rangpur
    num_bytes: 395864254.728
    num_examples: 1152
  - name: sylhet
    num_bytes: 395868121.728
    num_examples: 1152
  download_size: 3125125096
  dataset_size: 3560964103.5520005
---

# _Many Dialects, Many Languages, One Cultural Lens_: Evaluating Multilingual VLMs for Bengali Culture Understanding Across Historically Linked Languages and Regional Dialects

[![Website](https://img.shields.io/badge/Website-Click-blue)](https://labib1610.github.io/BanglaVerse)
[![Paper](https://img.shields.io/badge/Paper-Arxiv-red)](https://arxiv.org/abs/2603.21165v1) 
[![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Dataset-yellow)](https://huggingface.co/datasets/FaiyazAbdullah114708/BanglaVerse)

_**Abstract**_: Bangla culture is richly expressed through region, dialect, history, food, politics, media, and everyday visual life, yet it remains underrepresented in multimodal evaluation. To address this gap, we introduce BanglaVerse, a culturally grounded benchmark for evaluating multilingual vision–language models (VLMs) on Bengali culture across historically linked languages and regional dialects. Built from 1,152 manually curated images across nine domains, the benchmark supports visual question answering and captioning, and is expanded into four languages and five Bangla dialects, yielding ∼32.3K artifacts. Our experiments show that evaluating only standard Bangla overestimates true model capability: performance drops under dialectal variation, especially for caption generation, while historically linked languages such as Hindi and Urdu retain some cultural meaning but remain weaker for structured reasoning. Across domains, the main bottleneck is missing cultural knowledge rather than visual grounding alone, with knowledge-intensive categories. These findings position BanglaVerse as a more realistic test bed for measuring culturally grounded multimodal understanding under linguistic variation.

---

## Methodology
![Methodology](assets/methodology.png)
*Figure 1: Overview of the BanglaVerse dataset and experimental setup. The figure shows the two task types, example annotations for each task, artifacts generation and evaluation pipeline with multiple metrics.*

---

## Dataset Summary
The benchmark is built from **1,152 manually curated images** spanning nine culturally rich domains. It is expanded into four languages and five major Bangla dialects, resulting in approximately **32.3K total artifacts**.

### Core Statistics
* **Images**: 1,152 unique source images.
* **Tasks**: Supports Visual Question Answering (VQA) and Image Captioning (CAP).
* **Total Artifacts**: 10,377 captions and 20,727 VQA pairs (32,256 total).
* **Languages**: Bangla, English, Hindi, and Urdu.
* **Bangla Dialects**: Barishal, Chittagong, Noakhali, Rangpur, and Sylhet.

### Cultural Domains
The dataset encapsulates Bengali identity through nine carefully selected domains: _Culture, Food, History, Media & Movies, National Achievements, Nature, Personalities, Politics, and Sports_.

---

## Key Findings
* **Dialectal Sensitivity**: Evaluating only on standard Bangla overestimates model capability; performance drops consistently under dialectal variation, particularly for free-form caption generation.
* **Cross-lingual Preservation**: Historically linked languages like Hindi and Urdu preserve more cultural meaning in descriptive tasks than standard translation baselines suggest.
* **The Knowledge Bottleneck**: The primary limitation in understanding is missing **cultural knowledge** (e.g., named entities and specific context) rather than simple visual grounding or linguistic variation alone.

---

💻 All the codes, resources, and evaluation scripts are available on the [Project Website](https://labib1610.github.io/BanglaVerse).

💾 Downloading the Full Corpus

You can easily download and use the BANGLA VERSE dataset in multiple ways:

🔹 **Option 1: Using the 🤗 Datasets Library**

```python
from datasets import load_dataset

# Load the dataset from Hugging Face
dataset = load_dataset("FaiyazAbdullah114708/BanglaVerse")

# Check available splits
print(dataset)
```

🔹 **Option 2: Clone Directly from Hugging Face**

You can also clone the repository directly using Git LFS:

```bash
git lfs install
git clone https://huggingface.co/datasets/FaiyazAbdullah114708/BanglaVerse
```

🔹 **Option 3: Manual Download**

Visit the dataset page and use the “Download Dataset” button: 👉 https://huggingface.co/datasets/FaiyazAbdullah114708/BanglaVerse

---

## Citation
If you use this dataset or benchmark in your research, please cite:

```bibtex
@article{sayeedi2026many,
  title={Many Dialects, Many Languages, One Cultural Lens: Evaluating Multilingual VLMs for Bengali Culture Understanding Across Historically Linked Languages and Regional Dialects},
  author={Sayeedi, Nurul Labib and Sayeedi, Md Faiyaz Abdullah and Dipta, Shubhashis Roy and Tabassum, Rubaya and Hridoy, Ariful Ekraj and Mahmood, Mehraj and Sobhani, Mahbub E and Hasan, Md Tarek and Shatabda, Swakkhar},
  journal={arXiv preprint arXiv:2603.21165},
  year={2026}
}
```