Implementation:Open compass VLMEvalKit MedqbenchPairedDescriptionDataset

Field	Value
source	VLMEvalKit
domain	Vision, Benchmarking, Medical Imaging, Paired Description

Overview

Benchmark dataset implementation for MedQ-Bench paired medical image description evaluation in VLMEvalKit.

Description

MedqbenchPairedDescriptionDataset inherits from ImageBaseDataset and implements the MedQ-Bench Paired Description benchmark for evaluating paired medical image description capabilities. The TYPE field is set to 'Caption'. It supports dev and test splits and includes a MedQBench_PairedDescription_Scorer class for evaluating paired medical image descriptions.

Usage

Registered in vlmeval/dataset/__init__.py and invoked through build_dataset() by benchmark name.

Code Reference

Source: vlmeval/dataset/medqbench_paired_description.py, Lines: L1-445
Import: from vlmeval.dataset.medqbench_paired_description import MedqbenchPairedDescriptionDataset

Signature:

class MedqbenchPairedDescriptionDataset(ImageBaseDataset):
    TYPE = 'Caption'
    DATASET_URL = {...}
    DATASET_MD5 = {...}
    ...

I/O Contract

Direction	Description
Inputs	TSV dataset file with paired medical images and description tasks
Outputs	Evaluation results DataFrame with paired description scores

Usage Examples

from vlmeval.dataset import build_dataset
dataset = build_dataset('MedqbenchPairedDescription_dev')

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment