datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
InfoDet
InfoDet: A Dataset for Infographic Element Detection
🤗 Dataset | 🖥️ Code | 📄 Paper | 📄 Arxiv
🔥 News
[2026.2] 🎉🎉 InfoDet has been accepted by ICLR 2026!
[2025.11] 🎉🎉 We have released the first version of our dataset, which includes 11,264 real and 90,000 synthetic infographic charts, with 14,227,680 bounding box annotations.
✨ Dataset Description
InfoDet is a dataset designed to support the development of accurate object detection models… See the full description on the dataset page: https://huggingface.co/datasets/InfoDet/InfoDet.InfoChartQA
InfoChartQA: Benchmark for Multimodal Question Answering on Infographic Charts
🤗Dataset
Dataset
You can find our dataset on huggingface: 🤗InfoChartQA Dataset
Usage
Each question entry is arranged as follows. Note that for visual questions, there may be some extra input figures, which are cropped from the orginal figure. We present their bboxes in "extra_input_figure_bbox".
{
"question_id": id of the question,
"question_type_name":… See the full description on the dataset page: https://huggingface.co/datasets/Jietson/InfoChartQA.vsp-infoseek
InfoSeek Dataset (with images)
Knowledge-intensive visual question answering dataset.
Images sourced from OVEN,
metadata from InfoSeek.
infovqa_test_subsampled_beirBEIR version of vidore/infovqa_test_subsampled.
VisRAG-Ret-Test-InfoVQA
Dataset Description
This is a VQA dataset based on Infographics from InfoVQA dataset from InfoVQA.
Load the dataset
from datasets import load_dataset
import csv
def load_beir_qrels(qrels_file):
qrels = {}
with open(qrels_file) as f:
tsvreader = csv.DictReader(f, delimiter="\t")
for row in tsvreader:
qid = row["query-id"]
pid = row["corpus-id"]
rel = int(row["score"])
if qid in qrels:… See the full description on the dataset page: https://huggingface.co/datasets/openbmb/VisRAG-Ret-Test-InfoVQA.infovqa_test_subsampled_beirBEIR version of vidore/infovqa_test_subsampled.
Physics-Informed-Deformable-Gaussian-Splattingvlm-info-loss-results
VLM Grounding Evaluation Results
Grounding evaluation results for vision-language models on robotics manipulation datasets.
Part of the vlm-info-loss project studying
how VLM connectors transform visual representations.
Background
Our embedding-level analysis shows VLM connectors perform a compress-then-expand transformation:
they sharpen dominant-object representations while compressing secondary-object category identity.
All tested models converge to ~83%… See the full description on the dataset page: https://huggingface.co/datasets/MicroAGI-Labs/vlm-info-loss-results.OpenHumnoidActuatedFaceData
🤖 Open Humanoid Actuated Face Dataset
Dataset Summary
The Open Humanoid Actuated Face Dataset is designed for researchers working onfacial‑actuation control, robotics, reinforcement learning, and human–computer interaction.
Origin – collected during a reinforcement‑learning (RL) training loop whose objective was to reproduce human facial expressions.
Platform – a modified i2Head InMoov humanoid head with a silicone skin.
Control – 16 actuators driving facial… See the full description on the dataset page: https://huggingface.co/datasets/infosysLimited/OpenHumnoidActuatedFaceData.BToks-vidore_rag_Infographic-VQA
BToks ViDoRe Infographic-VQA
This dataset repository contains Lance-format converted data used by the open-source reproduction code for Bottleneck Tokens for Unified Multimodal Retrieval (arXiv:2604.11095).
Source
Converted from vidore/colpali_train_set.
Subset/view: infographic_vqa. This repository does not change upstream ownership, licensing, citation requirements, or usage restrictions.
Format
The data is stored as Lance tables for the… See the full description on the dataset page: https://huggingface.co/datasets/siyrus/BToks-vidore_rag_Infographic-VQA.FZI-AURA
FZI-AURA
A multimodal autonomous-driving dataset featuring the largest LiDAR sensor
suite of any public autonomous-driving dataset.
Download |
Python SDK |
Data format |
License
2,473 scenes |
13.70 hours |
8 cameras |
up to 12 LiDARs |
4.11 million 3D-box annotations |
30.01 billion segmented LiDAR points
Public preview release: 1,081 out of the 2,473 FZI-AURA scenes are currently available. The remaining scenes will be added soon.
Pin a… See the full description on the dataset page: https://huggingface.co/datasets/fzi-forschungszentrum-informatik/FZI-AURA.InfographicVQAInfographicsVQAI processed the jsonl file downloaded from DUE Benchmark using the code below.For the images, I used images provided by DocDownstream.
import json
from datasets import Dataset, DatasetDict
from tqdm import tqdm
from huggingface_hub import login; login(token="YOUR_TOKEN")
data = []
jsonl_file = "document.jsonl"
with open(jsonl_file, "r", encoding="utf-8") as f:
for line in tqdm(f):
record = json.loads(line)
img_path =… See the full description on the dataset page: https://huggingface.co/datasets/Ryoo72/InfographicsVQA.infovqa_test_subsampled
Dataset Description
This is the test set taken from the InfoVQA dataset. includes infographics collected from the Internet using the search query “infographics”. Questions and answers were manually annotated.
Questions and answers were manually annotated.
Example of data : (see viewer)
Data Curation
To ensure homogeneity across our benchmarked datasets, we subsampled the original test set to 500 pairs and renamed the different columns.
Load the dataset
from… See the full description on the dataset page: https://huggingface.co/datasets/vidore/infovqa_test_subsampled.BToks-visrag_indomain_InfoVQA
BToks VisRAG InfoVQA
This dataset repository contains Lance-format converted data used by the open-source reproduction code for Bottleneck Tokens for Unified Multimodal Retrieval (arXiv:2604.11095).
Source
Converted from openbmb/VisRAG-Ret-Train-In-domain-data.
Subset/view: InfoVQA. This repository does not change upstream ownership, licensing, citation requirements, or usage restrictions.
Format
The data is stored as Lance tables for the… See the full description on the dataset page: https://huggingface.co/datasets/siyrus/BToks-visrag_indomain_InfoVQA.ktb-ocr-dataset
PDFA OCR Dataset
Curated and Published by KREATIVE TIME BOX
This dataset contains document page images along with their corresponding OCR layout bounding box annotations derived from PDFA document extraction pipelines.
Dataset Overview
Organization / Creator: KREATIVE TIME BOX
Images: 27,499 PNG files (~8.8 GB)
JSON Annotations: 6,989 JSON files (~52 MB)
Image Format: PNG (RGB document page renders)
Annotation Format: JSON with text lines, normalized bounding… See the full description on the dataset page: https://huggingface.co/datasets/infokreativetimebox/ktb-ocr-dataset.InfoVQAinfovqa_beirThis is a copy of https://huggingface.co/datasets/jinaai/infovqa reformatted into the BEIR format. For any further information like license, please refer to the original dataset.
Disclaimer
This dataset may contain publicly available images or text data. All data is provided for research and educational purposes only. If you are the rights holder of any content and have concerns regarding intellectual property or copyright, please contact us at "support-data (at) jina.ai" for… See the full description on the dataset page: https://huggingface.co/datasets/jinaai/infovqa_beir.InfoSeekIT2ITRetrieval
InfoSeekIT2ITRetrieval
An MTEB dataset
Massive Text Embedding Benchmark
Retrieve source text and image information to answer questions about images.
Task category
it2it
Domains
Encyclopaedic
Reference
https://aclanthology.org/2023.emnlp-main.925
Source datasets:
MRBench/mbeir_infoseek_task8
How to evaluate on this task
You can evaluate an embedding model on this dataset using the following code:
import mteb
task =… See the full description on the dataset page: https://huggingface.co/datasets/mteb/InfoSeekIT2ITRetrieval.arabic_infographicsvqa_ar_beirThis is a copy of https://huggingface.co/datasets/jinaai/arabic_infographicsvqa_ar reformatted into the BEIR format. For any further information like license, please refer to the original dataset.
Disclaimer
This dataset may contain publicly available images or text data. All data is provided for research and educational purposes only. If you are the rights holder of any content and have concerns regarding intellectual property or copyright, please contact us at "support-data (at)… See the full description on the dataset page: https://huggingface.co/datasets/jinaai/arabic_infographicsvqa_ar_beir.informes_discriminacion_gitana
Resumen del dataset
Se trata de un dataset en español, extraído del centro de documentación de la Fundación Secretariado Gitano, en el que se presentan distintas situaciones discriminatorias acontecidas por el pueblo gitano. Puesto que el objetivo del modelo es crear un sistema de generación de actuaciones que permita minimizar el impacto de una situación discriminatoria, se hizo un scrappeo y se extrajeron todos los PDFs que contuvieron casos de discriminación con el formato… See the full description on the dataset page: https://huggingface.co/datasets/somosnlp-hackathon-2023/informes_discriminacion_gitana.infograph_train_10kinfovqa_trainmbeir_infoseek_task6infographic-VQAbizgen_infographic_bbox
BizGen: Advancing Article-level Visual Text Rendering for Infographics Generation (Glyph-ByT5-v3) (BBox Output, As layout Input)
Dataset Overview
This dataset supports research in article-level visual text rendering for business content generation, including infographics and presentation slides with ultra-dense layouts.
Key Features
📊 Layout Capabilities
• Supports ultra-dense layouts with 50+ layers
• Handles article-level… See the full description on the dataset page: https://huggingface.co/datasets/svjack/bizgen_infographic_bbox.erairaws_infosThis is the information integration of erai-raws and myanimelist.
Animes
2056 animes in total, 500 animes shown.
ID
Post
Bangumi
RSS
AniDB
AniList
Kitsu
LiveChart
MAL
Resources
Published At
Last Published At
60410
Yami Shibai 14
RSS
AniDB
AniList
Kitsu
LiveChart
MAL
20
2025-01-18T23:52:00
2025-03-09T14:25:00
60407
Kimi to Idol Precure
RSS
AniDB
Kitsu
LiveChart
MAL
12
2025-02-01T20:24:00
2025-03-09T14:39:00
60108
One Piece: Gyojin Tou-hen
RSS
AniDB
AniList
Kitsu… See the full description on the dataset page: https://huggingface.co/datasets/deepghs/erairaws_infos.infographicvqa_disco
InfographicVQA_mini Dataset
A mini subset of the InfographicVQA dataset with 500 randomly selected question-answer pairs for infographic visual question answering evaluation.
Dataset Details
Total Samples: 500 QA pairs
Source: InfographicVQA validation set
Task: Infographic Visual Question Answering
Image Format: PNG (extracted from parquet-embedded images)
Features: Includes pre-extracted OCR text from AWS Textract
Features
Each sample contains:… See the full description on the dataset page: https://huggingface.co/datasets/kenza-ily/infographicvqa_disco.douban_movie_info该数据集为豆瓣电影信息维表。
更多信息请参考文章《数据获取：豆瓣电影信息爬取》。
poster-schedule-information-extraction
Multimodal Visual-Text Dataset for Poster Schedule Information Extraction
A ready-to-train Indonesian document AI dataset combining pixels, OCR tokens, spatial layout, and BIO entity labels.
Overview
What it is
127 Indonesian seminar and religious-study event posters with multimodal token-level annotations
Primary task
Schedule information extraction as token classification
Modalities
Image + text + 2D spatial layout
Coordinates… See the full description on the dataset page: https://huggingface.co/datasets/Ibnuck/poster-schedule-information-extraction.