Implementation:Open compass VLMEvalKit VDC

Field	Value
source	VLMEvalKit
domain	Vision, Benchmarking, Video Description, Captioning

Overview

Benchmark dataset implementation for VDC (Video Detailed Captioning) evaluation in VLMEvalKit.

Description

VDC inherits from VideoBaseDataset and implements the VDC benchmark for evaluating video detailed captioning capabilities. The TYPE field is set to 'Video-VQA'. It supports multiple subset modes including 'all', 'breakpoint', 'short', and 'detailed' caption types, with configurable data limits for subsampling.

Usage

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

Code Reference

Source: vlmeval/dataset/vdc.py, Lines: L1-425
Import: from vlmeval.dataset.vdc import VDC

Signature:

class VDC(VideoBaseDataset):
    TYPE = 'Video-VQA'
    MD5 = ''
    ...

I/O Contract

Direction	Description
Inputs	TSV dataset file with video paths and captioning tasks
Outputs	Evaluation results DataFrame with captioning quality scores

Usage Examples

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

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment