Skip to content

Dataset · alibaba-multimodal-industrial-ai

Industrial product feature extraction benchmark

alibaba-multimodal-industrial-ai/IndustryBench-MIPU

Image and Chinese text dataset for evaluating MLLMs in multi-image attribute extraction.

What's inside

The benchmark from Alibaba contains 10,000 to 100,000 examples with images and Chinese text. It features real photographs of industrial products: specification tables, factory nameplates, and technical drawings. The task involves extracting structured attribute values distributed across multiple heterogeneous images.

What you can do with it

  • Evaluating multimodal LLMs in multi-image visual question answering tasks
  • Testing OCR and technical document/nameplate parsing systems
  • Developing pipelines for automatic population of industrial product catalogs
Tasks
Image → text, Visual question answering
Languages
Chinese
Size
10K–100K rows
Downloads this month
27.6K
Likes
8

How to load it

from datasets import load_dataset

ds = load_dataset("alibaba-multimodal-industrial-ai/IndustryBench-MIPU")