Implementation:Open compass VLMEvalKit ImageYORNDataset

Field	Value
source	VLMEvalKit
domain	Vision, Benchmarking, Yes/No QA

Overview

Benchmark dataset implementation for Yes/No QA evaluation in VLMEvalKit.

Description

ImageYORNDataset inherits from ImageBaseDataset and implements yes/no visual question answering evaluation. The TYPE field is set to 'Y/N'. It supports MME, HallusionBench, POPE, AMBER, and VSR-zeroshot benchmarks with judge-based evaluation.

Usage

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

Code Reference

Source: vlmeval/dataset/image_yorn.py, Lines: L1-106
Import: from vlmeval.dataset.image_yorn import ImageYORNDataset

Signature:

class ImageYORNDataset(ImageBaseDataset):
    TYPE = 'Y/N'
    DATASET_URL = {...}
    DATASET_MD5 = {...}
    ...

I/O Contract

Direction	Description
Inputs	TSV dataset file with image/video paths and questions
Outputs	Evaluation results DataFrame with scores per category

Usage Examples

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

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment