Dataset · SimulaMet
Medical dataset Kvasir-VQA-x1 for endoscopy
SimulaMet/Kvasir-VQA-x1
Multimodal dataset of 159,549 question-answer pairs for benchmarking in gastrointestinal endoscopy.
What's inside
Includes 159,549 new QA pairs and is designed for benchmarking medical visual question answering (MedVQA) in the field of gastrointestinal endoscopy. Used in the MediaEval Medico 2025 challenge. Intended for working with English images and text.
What you can do with it
- Training and validation of visual-question-answering models for medicine
- Benchmarking analysis systems for gastrointestinal endoscopy data
- Tasks
- Visual question answering, Image text → text
- Languages
- English
- Size
- 100K–1M rows
- Downloads this month
- 21.6K
- Likes
- 18
Models trained on it
- SimulaMet/Qwen2.5-VL-KvasirVQA-x1-ftVisual question answering 88
- SimulaMet/Qwen2.5-VL-Transf-KvasirVQA-x1-ftVisual question answering 22
- SimulaMet/MedGemma-KvasirVQA-x1-ftVisual question answering 22
- JustATalentedGuy/PMC-Kvasir-VQA-x1-lora_250918-1352Visual question answering 0
- sageofai/MEDVQA-GI-2026-Task1-VQA 0
- sageofai/MEDVQA-GI-2026-Task2-Explainability 0
- sageofai/Qwen25VL-MEDVQA-GI-S1-subtask1_2 0
- sageofai/Florence2-MEDVQA-GI-2026-Task1 0
How to load it
from datasets import load_dataset
ds = load_dataset("SimulaMet/Kvasir-VQA-x1")