imagesajimeno/PubTabNettable-recognitiondocument-aiocrcomputer-visionhtmlpubmedwebdatasetimage-to-textlayout-analysisbenchmark

PubTabNet — 568K Table Images with HTML Annotations

Name: PubTabNet — 568K Table Images with HTML Annotations
Creator: DataBazaar
Keywords: ajimeno/PubTabNet, table-recognition, document-ai, ocr, computer-vision, html, pubmed, webdataset, image-to-text, layout-analysis, benchmark

About this data

568K+ scientific table images paired with HTML structure annotations, extracted from PubMed Central Open Access articles. Standard benchmark for image-based table recognition and document AI.

Schema

Name	Type	Description
png	STRUCT(bytes BLOB, path VARCHAR)	PNG image file as binary blob with embedded path reference to source table image
__key__	VARCHAR	Unique identifier for the WebDataset sample linking image and HTML annotation
__url__	VARCHAR	Source URL or reference path to the original table image in the dataset repository

Sample Data

Preview a sample of the data before downloading.

Free

Open dataset

Quality: No ratings

0 downloads

Seller: DataBazaar

Agent? No sign-up needed →

For AI Agents

Via MCP Server

# 1. Add to your agent's MCP config (claude_desktop_config.json or similar):
{
  "mcpServers": {
    "databazaar": { "command": "npx", "args": ["databazaar-mcp"] }
  }
}

# 2. Your agent can then call:
search_datasets({ query: "PubTabNet — 568K Table Images " })
// Found: e82724e4-9b1e-47f5-bf62-c21e5982a2b2
get_download_url({ dataset_id: "e82724e4-9b1e-47f5-bf62-c21e5982a2b2" })  // free — no API key needed

Via REST API

# Free dataset — no API key required:
curl https://api.databazaar.io/datasets/e82724e4-9b1e-47f5-bf62-c21e5982a2b2/download-url