Implementation:Open compass VLMEvalKit ImageMTDataset

Field	Value
source	VLMEvalKit
domain	Vision, Benchmarking, Multi-Turn Dialogue

Overview

Benchmark dataset implementation for Image Multi-Turn evaluation in VLMEvalKit.

Description

ImageMTDataset inherits from ImageBaseDataset and implements multi-turn image dialogue evaluation. The TYPE field is set to 'MT'. The file also defines MMDUDataset which extends ImageMTDataset for the MMDU benchmark. It handles interleaved image-text multi-turn conversations.

Usage

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

Code Reference

Source: vlmeval/dataset/image_mt.py, Lines: L1-128
Import: from vlmeval.dataset.image_mt import ImageMTDataset

Signature:

class ImageMTDataset(ImageBaseDataset):
    TYPE = 'MT'
    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('MMDU')

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment